Document of The World Bank FOR OFFICIAL USE ONLY MICRFICEUCPY LNi. ()9, [Bk TyVpF: (PER Report No. 969 PROJECT PERFORMANCE AUDIT REPORT BURKINA FASO SECOND BOUGOURIBA, AGRICULTURAL DEVELOPMENT PROJECT (CREDIT 1097-BUR) BOUCLE DU MOUHOMS AGRICULTURAL DEVELOPMENT PROJECT (CREDIT 1284-BUR) HAUTS-BASSINS AGRICULTURAL DEVELOPMENT PROJECT (CREDIT 1285-BUR) KOUDOUGOU AGRICULTURAL PILOT PROJECT (CREDIT 1293-BUR) JUNE 28, 1991 Operations Evaluation Department This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. CURRENCY EQUIVALENTS French African Community Franc (CFAF) Pre-1982 US$1 - CFAF 300 1982-1987 = US$1 - CFAF 350 1988-ff = US$1 - CFAF 300 FISCAL YEAR January 1 - December 31 ABBREVIATIONS ADP = Agricultuial development project BOAD West African Development Bank CCCE = Central Economic Cooperative Bank CFAF = African Francophone Community Francs CFDT = French Textile Development Company CIRAD = Center for International Cooperation in Agronomic Research for Development CNCA = National Agricultural Credit Bank CRPA = Regional Agricultural Production Centers FAC = Aide and Cooperation Fund FAO Food and Agriculture Organization GTZ Technical Cooperative Corporation GV = Village Group IBRD = International Bank for Reconstruction and Development ICB International competitive bidding ICRISAT = International Crop Research Institute for the Semi-arid Tropics IDA - International Development Association IFAD = International Fund for Agricultural Development IRAT Tropical Agriculture Research Institute IRCT = Cotton Textile Research Institute IRR Internal rate of return M&E Monitoring and Evaluation NPfC Nitrogen, phosphate, potash OED Operations Evaluation Department OFNACER = National Ce,:eals Office ORD Regional Development Office PAPEM Pre-extension and Multi-site Experimentation support post PCR Project Completion Report PPA Project Performance Audit PPAR = Project Performance Audit Report PRESAP = Agricultural Support Services Project RD = Rural development SAR Staff Appraisal Report SMDR Cooperative Rural Development Society SOFITEX Burkinabe Textile Fiber Development Company SPN = Supervision T&V = Training/Vislt TOR = Terms of Reference UCECB = Burkina Union of Cooperative Savings Banks ULV Ultra low volume URCPB = Burkina Regional Union of People's Bank FMB OMVCUAL M49 OnY THE WOLD BANK WaShIngton. DC 20433 U.S A O1 Wo Dutw-Ceowal Opeeagm avauot"o June 28, 1991 MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECTs Project Performance Audit Report - Burkina Faso Second Bougouriba Agricultural Development Project (Credit 1097-BUR) Boucle du Mouhoun Agricultural Development Project (Credit 1284) Hauts-Bassins Agricultural Development Project (Credit 1284-BUR) Koudougou Agricultural Pilot Proiect (Credit 1293-BUR) Attached, for your information, is a copy of a report entitled "Project Performance Audit Report on Burkina Faso - Second Bougouriba Agricultural Development Project (Credit 1097-BUR), Boucle du Mouhoun Agricultural Development Project (Credit 1284), Hauts-Bassins Agricultural Development Project (Credit 1284-BUR), and Koudougou Agricultural Pilot Project (Credit 1293-BUR)", prepared by the Operations Evaluation Department. Attachment This document has a restricted distribution and may be used by recipients only in the Performance of their oficial duties. Its contents may not otherwise be disclosed without World Bank authorisation. FOR OFFICIAL USE ONLY PROJECT PERFORMANCE AUDIT REPORT BURKINA FASO SECOND BOUGOURIBA AGRICULTURAL DEVELOPMENT PROJECT (CR. 1097-BUR) BOUCLE DU MOUHOUN AGRICULTURAL DEVELOPMENT PROJECT (CR. 1284-BUR) HAUTS-BASSINS AGRICULTURAL DEVELOPMENT PROJECT (CR. 1285-BUR) KOUDOUGOU AGRICULTURAL PILOT PROJECT (CR. 1293-BUR) TABLE OF CONTENTS Paze No. Preface. . . . . . . . . . . * * * * * * . . . . i Basic Data Sheet . . . . . . . . . * * * * * * * * * * * * * . iii Evaluation Summary . . . . . . . . . . . . . . . . . . . . . . . xiii Project Performance Audit Report I. Introduction . . . . . . . . . . . . . . . . . . . . . 1 II. Background . . . ........******** ***2 III. Project Design . . . . . . . . . . . . . . . . . . . . 5 IV. Implementation . . . . . . . . . . . . . . 11 V. Impact . . . . . . . . . * * * * * * * * * . 17 VI. Issues and Findings . . . . . . . * * * * * * * * . . 21 A. Ecological effect . . . . . . . . * * * * * * * * * 21 B. The Koudougou downgrade ............. 21 C. Extension . . . . . * . . . . * * * * * * * * . . . 23 D. Area-based or country-wide . . . . . . . . . . . . 26 E. aral credft . . . . . . . . . * * * * * * * * * * 27 Attachment: Comments from Cofinancier . . . . . . . . . . . . . 31 MAPS: IBRD 15915R IBRD 15127R1 IBRD 14815R IBRD 14816R IBRD 15914R1 This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authoriza on. PROJECT PERFORMANCE AUDIT REPORT BURKINA rASO SECOND BOUGOURIBA AGRICULTURAL DEVLOPMENT PROJECT (CR. 1097-BUR BOUCLE DU MOUHOUN AGRICULTURAL DEVELOPMENT PROJECT (CR. 1284-BUR)I HAUTS-BASSINS AGRICULTURAL DEVELOPMENT PROJECT (CR. 1285-BUR) KOUDOUGOU AGRICULTURAL PILOT PROJECT (CR.1293-BUR) PREFACE 1. This is the Project Performance Audit Report (PPAR) for four similar agricultural development projects in vestern Burkina Faso (formerly Upper Volta). All four were area-based projects. Each was executed by a Regional Development Office (ORD). During project implementation, these government bodies were abolished and, after a hiatus of more than a year, re-created at Regional Agricultural Production Centers (CRPA), with somewhat different functions. The four were as follow: Prolect Name Cred # Cr. Amount Approval Closing Final Disb. (US$ mill> (date> (date> (date> Bougouriba II 1097 16.0 28 Jan'81 31 Mar'89 27 Dec'89 Boucle du 1284 6.8 27 Jul'82 30 Jun'88 23 Jan'89 Mouhoun Hauts Bassins 1285 4.7 27 Jul'82 30 Jun'88 2 Jan'89 Koudougou Pilot 1293 7.0 21 Sep'82 31 Mar'89 27 Oct'89 Substantial amounts of each credit were cancelled as follows: US$3.73 million (23%) of Bougouriba II; $4.25 million (62.5%) of Boucle du Mouhoun; $2.9 million (62%) of Hauts-Bassins; and $2.05 million (292) of Koudougou Pilot. As of 30 September 1990, no debt service payments were due from the Borrower. 2. The PPAR is based on the Project Completion Reports (PCRs), on the Staff Appraisal Reports for Bougouriba II, Boucle du Mouhoun and Hauts- Bassina, on the draft staff appraisal report and preparation report for Koudougou Agricultural Development Project (the precursor of Koudougou Pilot), on the President's Reports and legal agreements for all four projects, on the transcripts of Executive Directors' meetings at which the projects were considered, on study of project files, and on discussions with IDA staff. 1 Called Volta Noire at time of credit approval and until 1983. 3. An OED mission viaited Burkina Faso in November 1990. The mission discussed the effectiveness of the four projects with Burkinabe Government officials of the Ministry of Agriculture and Livestock, including staff of all project CRPAs, Ministry of Planning, Ministry of National Education, Ministry of Territorial Administration, and of the following cooperative or parastatal organizations: the National Agricultural Credit Bank (CNCA), Bougouriba Regional Union of People's Banks (URCPB), Burkinabe Textile Fiber Company (SOFITEX), and the National Cereals Office (OFNACER), as well as with farmers and ranchers of the project areas and other private citizens. The mission also discussed the four projects with Ouagadougou representatives the Dutch Ministry of Cooperation and the French Central Economic Cooperation Bank (CCCE), co-financiers respectively of Koudougou Pilot, and of Boucle du Mouboun & Haute Bassins, as well as with representatives of the European Development Fund. The mission visited Paris for further discussions with CCCE, which had already issued its own PCR and retrospective evaluation on Boucle du Mouhoun and Haute Bassins. 4. Except as noted in the Project Performance Audit, the PCRs provide a reasonably accurate account of projects' experience, including ptrformance of IDA, of the Borrower, including the executing ORDs/CRPAs, and of other participants in project execution except for the cofinanciers. The PPAR explores certain aspects of projects' design and implementation in greater depth. It reviews the debates between technology and environmental optimists and pessimists which shaped the downgrading the the Koudougou project, the genesis of the Village Land Management (Gestion des Terroirs Villageoises) project, and tensions between IDA and certain cofinanciers. It explores the evolution of the Borrower's and IDA's conception of agricultural extension from the T&V approach to the National System, and tensions with cofinanciers who want the evolution to go further. It compares the strategy of promoting rural development through area projects, as in these four projects, with the country-wide strategy that has supplanted it. 5. Following standard OED procedures, copies of the draft PPAR were sent to the Borrower and its executing agencies, and to cofinanciers. Comments received from the Caise Centrale de Cooperation Economique are reproduced as an Attachment to the PPAR. PROJECT PERFORMANCE AUDIT REPORT BURKINA FASO SECOND BOUGOURIBA AGRICULTURAL DEVELOPMENT PROJECT (CREDIT 1097-BUR) BASIC DATA SHEET Key Project Data Original Actual or Plan Current Estimate Project Cost (US$ million) 17.5 12.5 Credit Amount (US$ million) 16.0 16.0 - Disbursed - 10.98 - Cancelled 3.73 vate for Completion of Physical March 31, 1986 Sept. 30, 1988 Components Proportion Completed by Target Date (%) 100 100 Economic Rate of Return (%) 23 18 Institutional Performance - Satisfactory Project Timetable Phase Original date Revised date Actual Date Identification 1977 Preparation January 1979 Appraisal mission February 1980 Negotiations November 1980 Approval Jan. 29, 1981 Signature March 6, 1981 Effectiveness June 15, 1981 December 4, 1981 Project Completion March 31, 1986 March 31, 1987 Sept. 30, 1988 Credit closing date Sept. 30, 1986 Sept. 30, 1987 March 31, 1989 Borrower Republic of Burkina Faso Executing Agency Bougouriba Regional Development Authority FY of Borrower Calendar Year Follow-up Project None - iv - Project Costa and Financing A. Project Costs Est Act &LE Est Act A/K --(US$ m)-* (M) --(CFAF Ma)-- (2) Agricultural Extension 4.9 4.0 81 1030 1383 134 Training and Extension Methods 0.9 0.7 77 184 247 134 Supply Service 2.1 1.0 48 441 334 76 Oxen, Equipment, Inputs 3.7 2.3 62 772 822 106 Livestock 1.4 0.7 50 294 257 97 Tree Planting 0.4 0.3 75 74 95 128 Headquarters 3.7 3.2 86 772 1129 146 Monitoring and Studies 0.5 0.3 60 110 97 88 Total 17.5 12.5 71 3678 4364 119 Notes From CFAF 210/$1 at appraisal, the average rate over the disbursement period rose to CAF 350. Thus while project costs in local currency increased by 202, project costs in dollars fell by 302. B. Project Financing Est Act A/ Est Act A --(US$ m)-- (2) --(CEAF m)-- (2) IDA 16.0 10.1 63 3360 3542 105 Government 1.5 2.4 160 315 822 261 Total 17.5 12.5 71 3675 4364 119 Cumulative Diabursements FY81 FY82 FY88 FYe4 FY8-B FY86 FY87 FY881 FY89 FYN9 Appis.al Estimate (USS milon) .*0 9.7 1.0 1..7 1616.05 1 Apprta IUst mtliN lII n - 1.14 2.85 4.20 5.68 6.52 7.76 8.89 166 16089 Actual (US$ mill Ion) e.0 81.7 41.9 44.2 88.7 41.5 48.1 55.6 62.8 68.6 Actual as X of Es6tmat* -v * Use of Bank Resources A. Staff Inputs FY78 79 80 81 82 83 84 s 86 67 66 89 90 Total Preparation 16.3 10.3 6.2 32.8 Appraisal 53.2 13.7 669 Negotiations 6.6 6.6 Supervision 10.2 10.4 5.9 11.0 6.3 8.2 2.9 5.5 6.6 66.9 Other 24.1 11.0 5.5 40.6 Total 16.3 10.3 83.4 41.4 10.4 5.9 11.0 6.3 8.2 2.9 5.5 6.6 5.5 213.8 B. issions (1) Phase of month/ No. of No. of Specialists Project Types project cycle Year people drys represented perfor- of (2) mance problems ** ***------------------------------------------------------------ Identification 12177 1 unknown Preparation 4-5179 12 5 asc,d,e Appraisal 2.3/80 unknown Evaluation 5/80 4 19 a,c,d,e Supervision 1 3/81 3 6 a,d 1 F 2 1/81 2 6 a,d 2 FIN 3 6/81 2 7 a,d 2 M,0 4 1/83 2 5 a,d 1 M 5 6/83 1 4 a 2 M,T 6 10/83 1 4 a 2 F 7 2184 1 unspecified a - 8 5/84 1 3 a 2 F,M.P. 9 2/85 1 4 a 2 H,P 10 7-8/65 unspecified 2 * 12 6-7/87 3 8 a,c 2 - 13 4/88 2 5 a 2 - 14 11/88 1 4 £ 2 (1) Staff and Consultants (2) Specialized fields as agetimist, b- forestry expert. ca livestock ezpert, d- financial analyst, *o econmist, agricultural-economist, fa public works engineer go operations officer from the Resident ission. - vi - PROJECT PERFORMANCE AUDIT REPORT BURKINA FASO BOUCLE DU hNOUHCUN AGRICULTURAL DEVELOPMENT PROJECT (CREDIT 1284-BUR) BASIC DATA SHEET Key Prolect Data Original Actual or Plan Current Estimate Project Cost (US$ million) 22.1 14.9 Credit Amount (US$ million) 6.8 6.8 - Disbursed - 2.7 - Cancelled 4.2 Cofinancing: - IFAD 7.9 3.1 - France 3.1 N/A - Comm=ercial Banks 3.5 N/A - Beneficiaries 0.8 N/A Date for Completion of Physical June 30, 1987 June 30, 1988 Components Proportion Completed by Target Date (1) 100 60 Economic Rate of Return (2) 21 22 Institutional Performance - Moderate Cuwul&tivegstiMated and A ntU& nDi1hUaUan§= Appraisal Estimate (US$ M) 0.3 1.6 3.3 5.0 6.3 6.8 - Actual (US$ a) 0.0 0.6 0.8 1.0 1.6 2.3 2.7 Actual as % Estimate 0 38 24 20 25 34 40 szaff .IRn nt s= EM E= m= M EM Appraisal 13.9 13.9 Negotiations 3.1 0.7 3.8 Supervision 5.0 8.0 8.2 7.0 3.4 13.2 1.2 46.0 Completion 3.0 0.4 3.4 TOTAL 17.0 5.7 8.0 8.2 7.0 3.4 13.2 4.2 0.4 67.1 - VAi - Other Proiect Data imS WgainAl Plan bs'isins Actuil First Mention in Bar.k Files 10/79 - 10/79 Negotiations 9/81 1/82 4/82 Board Approval 3/82 7/82 Effectiveness 9/82 12/82 6/83 Closing Date 6/30/J8 - 6/30/88 Borrower Republic of Burkina Faso Executing Agency Boucle du Mouhoun Regional Development Authority FY of Borrower Calendar Year Follow-up Project Agricultural Support Services Project (Credit 1979-BUR) Mission Data Icem Month/ No. of No. of Staff Discipline Date of - WIe kso Persons es Reort Preparation A/ 7/81 Appraisal ]/ 10/81 2 7 14 FA, EC, AGRON 6/28/82 Supervision / 1 1/83 1 2 2 AGRON 3/08/83 2 11/83 1 2 2 AGRON 3/06/84 3 6/84 1 2 2 AGRON 11/20/84 4 11/84 1 2 2 AGRON 3/06/85 5 6/85 1 2 2 AGRON 9/11/85 6 12/85 1 2 2 AGRON 1/22/86 7 5/86 1 2 2 AGRON 6/25/86 8 1/87 1 2 2 AGRON 2/20/87 9 6/87 2 2 4 SOCIOL, AGRON 7/20/87 10 4/88 1 2 2 AG-ECON 6/24/88 A/ Done by a consultant team financed by FAC/CCCE; detailed mission data not readily available. h/ Appraised in conjunction with the Hauts-Bassins Agricultural Development Project (Credit 1285-BUR); one-half of mission time allocated to each project. Four of the seven appraisal mission meabers were from cofinanciers FAC, CCCE and IFAD; their time is not included in the staff Inputs table above. g/ All supervision missions vere conducted jointly with CCCE (whose time is not included in the stafi inputs table above). Additional, partial supervisions of the Livestock component of the project were carried out in conjunction with supervisions of the Livestock Project; these partial supervisions did not generate separate full supervision reports for the Boucle du Mouhoun project. Exchane Rtes Name of Currency French African Community Franc (CFAF) Appraisal Year Average (1981) US$1 - CFAF 300 Intervening Years Average (1982-7) US$1 - CFAF 350 Completion Year Average (1988) US$1 - CFAF 300 - viii - PROJECT PERFORMANCE AUDIT REPORT BURKINA PASO HAUTS-BASINS AGRICULTURAL DEVELOPMENT PROJECT (CREDIT 1285-BUR) BASIC DATA SHEET Key Proiect Data Actual or Original P1au Currnt Estimate Project Cost (US$m) 20.8 11.4 Credit Amount (US$a) 4.7 4.7 Disbursed 1.9 Cancelled 3.4 Cofinancing: IFAD 3.1 1.9 Beneficiaries 0.5 0.3 .ommercial Banks 3.2 3.7 France 9.3 2.7 Government 0.0 0.9 Date for Completion Physical Components 6/30/87 6/30/88 Proportion Completed by Target Date (0) 100 60 Economic Rate of Return (0) 33 20 Institutional Performance Noderate Cumulative Estimated and Actual DisburAement Appraisal Estimate (US$a) 0.2 1.1 2.3 3.4 4.7 - Actual (US$m) 0.0 0.5 0.8 0.9 1.1 1.6 1.9 Actual as 0 Estimate 0 45 35 26 23 34 40 Staff_Inont ZX92 EM33 FY84 M3-.=5 EMS Y7 MA3 Mg3 FY0 TOTAL Appraisal 14.0 14.0 Negotiations 3.1 0.9 4.0 Supervision 4.9 7.9 10.6 6.9 3.0 15.8 2.7 0.4 52.3 Completion 1.0 1.9 0.5 3.4 TOTAL 17.1 5.8 7.9 10.6 6.9 3.1 16.8 4.6 0.9 73.7 -ix - Other Proleac Data 128Original Plan Revisionst wuda First Mention in Bank Files 10/79 - 10/79 Negotiations 9/81 1/82 4/82 Board Approval 3/82 - 7/82 Effectiveness 9/82 12/82 6/83 Closing Date 6/30/88 6/30/88 Borrower Republic of Burkina Faso Executing Agency Haut Bassins Regional Development Authority FT of Borrover Calendar Year Follow*up Project Agricultural Support Services Project (Credit 1979 - Bur) Item Moth/ No. of No. of Staff Discipline Date of - ells Persons ZSk Rgr Preparation A/ 7/81 Appraisal b/ 10/81 2 7 14 EA,KC,AGRO 6/28/82 Supervision /1 1/83 1 2 2 AGRO 3/08/83 2 11/83 1 2 2 AGRO 3/06/84 3 6/84 1 2 2 AGRO 11/26/84 4 11/84 1 2 2 A0 3/06/85 5 6/85 1 2 2 AGRO 9/11/85 6 12/85 1 2 2 AGRO 1/22/86 7 5/86 1 2 2 AGRO 6/25/86 8 1/87 1 2 2 AGRO 2/20/87 9 6/87 2 2 4 SOCIOL,AGRO 7/25/87 10 4/88 1 2 2 AG-ION 6/24/88 A/ Done by a consultant team financed by FAC/CCCE; detailed mission dta not readily available. b/ Appraised in conjunction with the Volts Noire Agricultural Development Project (Credit 1284-UV); one-half of mission time allocated to each project. Four of the seven appraisal mission members were from cofinanciers FAC, CCCE and IFAD; their time is not included in the staff inputs table above. S/ Al' apervision missions were conducted jointly with CCCE (whose time is not included in the staff inputs table above). Additional, partial supervisions of the livestock component of the project were carried out in conjunction with supervisions of the concurrent Livestock Project; these partial supervisions did not generate separate full supervision reports for the Hauts Bassins project. - x - PROJECT MRORMANCE AUDIT REPORT BURKINA FASO KOUDOUGOU AGRICULTURAL PILOT PROJECT (CREDIT 1293-BUR) BASIC DATA SHEET Key Project Data Original Actual or Plan Current Estimate Project Cost (US$ million) 10.8 6.6aI Credit Amount (US$ million) 7.0 7. - Disbursed -5.4 - Cancelled -2.05 Cofinancing (US$ million)* - Netherlands 2.9 Date for Completion of Physical March 31, 1988 Sept. 30, 1988 Components Proportion Completed by Target Date uncertain; design changed Economic Rate of Return (1) 241 6-8 Institutional Performance - Poor Proiect Timetable Phase Orizinal date Actual Date Identification October 19771/ October 1981 Preparation 19781/ Appraisal mission June 197911 October 1981 Negotiations November 1980yI June 1982 Approval September 21, 1982 Signature March 17, 1983 Effectiveness October 6, 1983 Project Completion March 31, 1984 September 30, 1988 Credit closing date September 30, 1988 March 31, 1989 Borrower Republic of Burkina Faso Executing Agency Kodougou Regional Development Authority FY of Borrower Calendar Year Follow-up Project None ---------------------------- a These dates are actua and refer to the Koudougoa Agricultural Development Project, which was abandoned because of uncertainties altout projected yield increases. a Although the IDA credit is closed, parts of the project continues with Dutch and Burkinabi financing. b/ Actual Netherlands financing will have been higher than shown here to the extent of any costs incurred directly by the Netherlands, or of 1 Ts 4e dai- A fte tne closin of the IDA credit. -xi- Preteet Cents and Finannt A. ProfetscoaMe A .-(US)--- () (CrAF al- (M) xensaion Servie 1.5 2.1 140 450 737 164 aputs and Equip.mnc h/ 2.1 0.9 43 630 315 50 Adaptve REsaarc 1.5 1.2 80 450 419 93 Aimal alth 9/ 1.2 0.2 13 360 53 15 .aoiton and Evaluacton 5/ 1.8 0.3 16 540 99 18 .aage~n and Adata. 2.7 2.0 74 810 699 86 :otal 10.8 6.6 61 3240 2322 72 m CYAF 300/31 a appraisal. thm .avør& rate over the dtaburnemsnt partod rose co CFAF 350. Thun the A/& % 1 lover t dollras. / Actual expe ures not known vith any precioion since inputs and equtpmnt vere finanned by borro ers and banks coo, but their costa vere not intagrally refleocud in projeot aceo~cs. S/ Financed in parallel by the Nøtbrland*. Åtua costs shown are the øxpendiVues recorded in projecc accounts and do not. thørefore. refleet any costa direecly ineurred by th Nethørlands. 5. Projaee Finaangn la. åa. åa ga. Au. åa --(u~$)-- (M -(CFAF M)- (t) IA 7.0 5.3 76 2100 1855 88 Netherland / 2.9 0.4 14 870 152 17 Beneficiaies 0.3 0.3 86 90 90 100 Coseretal,anks 0.3 0.3 86 90 90 100. Governmne 0.3 0.3 86 90 90 100 Total 10.8 6.6 61 3240 2322 72 A/ Acual MÅfinaming in dollas was $1.7 million laa than the appraisal esel~at (as detesmined from projeet aeounce kopt tn MFF). At the closing date, he amuc canelled vas equ~valeno to $2.4 mllion (as detetmind frm the Credit aomn kept in SD). kl Actual Ne~beriands finanoing vill have been highr than sho~. ber to the ~ssen of any cosca direetly incured by the Netherlads, or of costa raobured atter the closing of the IDA credit. CradAt efiIabumarane Cumulative Esnciaed and Aetual Dimhurmementa M83 -f4 Ms5 F6 FY7 Må Mi9 ...........---. -(SDI million)------------------- Appraisai EsUiat 0.5 1.5 * 2.7 4.1 5.2 5.9 6.3 ACtual 0.4 0.8 1.3 1.6 2.2 2.9 4.5 A (t) 80 53 48 39 42 49 71 Datoe of F~na DIsbursemmc: 10/27/89 S6aff Inmuta FYUi MYS FY64 "Ø C.O FWS FYØ9 FY89 F%q TOUAL Preparaton 4.8 4.8 Appraisul 48.1 4g.1 Nsgegio6 i.a 1.1 4.2 6.3 Supervien 7.8 6.2 19.6 8.1 8.4 7.6 2.2 63.4 Other 4.9 2.6 6.9 TOTAL 68.9 6.L 7.6 6.2 1e.8 8.5 8.4 7.6 2.2 116.6 Stage of Project Month/ No. of No. of Specialti8s2 Performance3 Types of Cycle Year Persons Days Probem4 IdentificatioD * Preparation * a;c;G Appraisal5 10/81 *;c;d;& Post-appraisal 1-2/82 2 l8 a;e Supervision 1 6/83 1 6 a 2 2 7/83 1 10 a 2 3 11/83 1 2 a 2 4 3/84 3 4 a,e i Not knówn 5 7/84 I 2 c .6 6 2/85 1 11 a 7 11/857 3 9 a;e 2 8 8-9/86 2 23 a;e 2 9 4/87 2 7 a;e 2 10 11/87 4 12 a;d;e i 11 4/88 unknown 2 12 5/89 1 4 g 13 12/89 '1 4 g * The project was firat identlfled in October 1977, appraied in June 1979 and negotiated in November 198, then reformulated es a p;lot project. l. Staff and consultantg 2. Specialista: (a) agronomiet; (b) foreatry apeciallet; (c) animal huebandry ope«lailst; (d) financl1 analyst; ;(e) econelset/agricultural econost; (f) public worke engineer; (g) operations officer, resident alesien. 3. Performanco: Up> to June 8s 19815: 1. No probløe, or minor probleme. 2. Intermedlate problems. 3. Important problems. From july l lø9s i. No problems, r minor problems. 2. Intermdlate probleme. 3. Important problems, satefactorily addressed. 4. Important problema, not satiefactorily addressed and requiring more attention from the borrower/execulng agency. 4. Key Problems: (F) financial; (M) management; (T) technicel; (P) poltical; (0) other. 6. Including Agrar and HydrotechnIk, Essen, Federal Republic of Germany consultante). 6. This special oupervision alssion was concørfed oxclu91vely with AVY and the extension service,. The over all statue of the Projeat vas not evaluated. 7. The ne fore intreduced In July 1986 doe not cell for typee of problem to be epeelf led. - xiii - PROJECT PERFORMANCE AUDIT REPORT BURKINA PASO SECOND BOUGOURIBA AGRICULTURAL DEVELOPMENT PROJECT (CR. 1097-1UR) BOUCLE DU MOUHOUN - AGRICULTURAL DEVELOPMENT PROJECT (CR. 1284-BUR) / HAUTS-BASSINS AGRICULTURAL DEVELOPMENT PROJECT (CR. 1285-BUR) KOUDOUGOU AGRICULTURAL PILOT PROJECT (CR.1293-BUR) EVALUATION SUMMARY Background that the best development course is to extend the improvements we are 1. These four projects mark a sure of while looking for solutions middle phase in IDA attempts to help to the remaining problems -- with the rural Burkinabe improve their lot. pausimiets one -- that what IDA and An earlier phase concentrated more on others had been pushing is agricultural production, especially ecologically damaging and that on converting farmers from hoe finding better recipes should take cultivation to ox-plowing, and on priority over extension. since introducing cottos into the bush- extension is a key element in these fallow rotation. The later, present projects, the Audit examines how phase stresses country-wide, service received doctrines have been adapted projects that exhibit technological to Burkinabe conditions in these and ecological pessimism in the sense projects and the current controversy that they seek wolutions to problems on the subject. The Audit weighs the of soil-fertility sustainability wisdcm of moving from an area-based which were previously thought to have ural developmnt strategy to a been solved, or, at least, to be country-base one. manageable. abiecties 2. In the middle phase then, the one under audit, the projects, while 4. All four projects aimed at still area-based, and while still improving the well-being of the rural essentially based on promoting ox- people in their areas. All of the plowing and cotton, were much broader four project areas are overwhelmingly in their concerns for the ecosystem rural and poor. Agricultural and for promoting its sustainable production is mainly for home exploitation. The credits for these consumption. In most of the area, four projects, for US$34.5 million, sorghum, millet, cowpeas, groundnuts were approved in 1981 and 1982; they and Bambara groundnuts are the main disbursed through 1989. At that crops, plus cotton, which is grown point, $12.93 million of undisbursed for sale and export. In the wettest, balances, 38% of the credit amounts, most-southerly part, maize and yams was cancelled. are more importantl in the driest, most-northerly part, millet is. 3. The Audit explores how these Drought makes cotton risky in the projects balanced the technical and north; humidity and associated ecological optimists, pouition -- diseases make cotton risky in the - xiv - south. Fulani pastoralists range subdivisions of the original project throughout the four-projects' area, area. and most cattle, goats and sheep are either owned by these specialists or, Implementation Experience if owned by farmers, are entrusted to them during part of the year. Farm- 7. Implementation of all four livestock integration is rare, but projects proved to be difficult. AS farm-livestock interactions are many. part of the Sankars,. revolution Transhumance to dry-season pastures (August 1983-October 1987), the ORDs, has proved more efficient than the project implementing growing and storing fodder. organizations, were abolished. They were in limbo for over a year before 5. The projects sought to they were re-created as CRPAs with persuade people to change their different functions. ORDs had farming and stock-raising practices, covered part of their extension costs promoting ox-plowing, farm-livestock with the money they raized selling integration, better use of animal and farm inputs, which is not a CRPA plant wastes, improved veterinary function. Not only has this made care, anti-erosive stone contour CRPAs completely financially bunds, tree planting, use of better dependent on Government or on aid seed ana purchased fertilizer, and programs, but extensionists complain other practices. The means the that, now that they have lost their projects used to persuade people were influence over input and credit improved extension, credit, applied allocation, farmers are less research and village organization, interested in their extension including literacy training. messages. Moreover, during the revolutionary period, the new 6. These objectives were accountability of civil servants to translated into specific targets in People's Revolutionary Tribunals did terms of outputst tons of grain, not encourage risk-taking and cotton; and of inputs: village groups initiative by the projects, managers. formed, extension visits, cattle Finally, administrative re- inoculated, research trials carried organization, notably, the setting up out. Since the proposed changes and of the political Ministry of the the means of persuasion were not new, Peasant Question, responsible for most gain was expected from some things traditionally under persuading more farmers to do what Agriculture and Livestock, some progressive farmers had done complicated implementation. before the projects started. Three of the projectat Bougouriba II, 8. Kaudougou Pilot was cofinanced Boucle du Mouhoun and Haute Bassins, with the Dutch; Boucle du Mouhoun and were substantially alike in this Haute Bassins were cofinanced with regard. The original Koudougou IFAD and with the French. project negotiated in November 1980 implementation revealed difficulties was similar. However, Koudougou between IDA and the Dutch, and Pilot, into which it was changed, was between IDA and the French. These more pessimistic about the are not covered in the PCR9, nor in traditional development formula and CCCE's audit of Boucle du Mouhoun and set out to discover useful changep Haute Bassins. 2 farmers and pastoralists could make through applied research in two small 9. Regarding oudougou, IDA's withdrawal from the negotiated - xv - project and the down-grade to the projects reported in the PCR9, Pilot represented a major change and confirming earlier favorable findings one not always handled with complete on Boucle du Mouhoun and Haute transparency with the cofinancier and Bassins by an OED Impact Evaluation Borrower. The Koudougou Pilot Studya/ and the recent ones of CCCE's conception, not terribly coherent in ex-post audit mentioned above. To the Audit's opinion, evolved further review, Boucle du Mouhoun and Haute during implementation. The lack of Bassins' implementation periods were adequate communication led to poor accompanied by large increases in sequencing of support by IDA and the cottin production and increases in Dutch for components in the reduced grain production. Their PCRs project. carefully net out increases stemming from in-migration which occurred 10. Regarding Boucle du Mouhoun independently of the projects. The and H"uts Bassins, there were remaining increases are enough to differences between the cofinanciers give an economic IRR of 20% for Haute over extension philosophy. These Bassins and of 35% for Boucle du came to a head during a joint mission Mouhoun. Bougouriba II claims an in October 1985 during which an IDA economic IRR of 18%, based representative emphasized the principally on increased grain importance of form and a CCCE production; there is no netting out representative emphasized the need of in-migration effects, but there for more appropriate content if was less in-migration than in other formal structures were to be project areats. Koudougou Pilot does worthwhile. More serious for the two not attempt to estimate any output projects were complications over effects but claims that its applied procurement. Misunderstandings over research came up with interesting IDA-IFAD insistence on international results. None of the projects claim competitive bidding and French success in farm-livestock insistence on tied procurement that integration. Livestock interventions should have been easy to anticipate were effectively limited to led to major delays and hard veterinary programs. Bougouriba 11 feelings. The SAR financing and Boucle du Mouhoun had some provisions seems unnecessarily success in persuading farmers to complex and not sufficiently build stone contour barriers. explicit. 13. More recent production data 11. All four projects were closed provided to the Audit by Bougouriba before their entire IDA credits could 11 show that cotton production has be spent. Partly as a result of dropped in Poni, but do not call for their implementation problems, they any change in PCR conclusions. After were closed out to make way for a visiting one of Koudougou Pilot's two different, country-wide strategy of applied research centers (PAPEMs), supporting agricultural development. the Audit is not so enthusiastic Bougouriba III was prepared without about the good sense and IDA support or encouragement and was applicability of work done there as not reviewed by IDA. the tCR is. Results 14. The Audit is impressed, as CCCE's audit is, that almost all 12. The Audit has little to add production increases cama from to the favorable results of these increases in area cultivated -- what - xvi - CCCE's audit calls extensification. set by population density, There are Yields increased little, and most of clearly unresolved problems about that increase resulted from increased farming and soil degradation in the urea consumption. If the area. Where population density is "quantitative" objectives were often low and fallows are long enough, they exceeded, what CCCE calls the are either short-term problems "qualitative" ones -- farm-livestock resolved by fallowing or they are integration, extension for herders, minor. And the projects did not literacy, increased use of organic tackle them in higher population manure -- were not exceeded. For areas where fallows are too short, these aspects, there is little to like the north of Roudougou. In the show. Audit's opinion, this is where it is most important to tackle them, and 15. So the verdict of success is also where intensification has the built on extensification -- expansion highest pay-off and is most likely to of the farming system previously occur. adopted by progressive farmers in the area, with increased ox-plowing, Findincs insertion of cotton into the cycle, and modest increase in fertilizer and 17. The Audit finds that soil pesticide use (all controlled for the degradation and other negative effect of in-migration). To what ecological trends stem largely from extent is that phenomenon a secular increase in population density and trend that would have occurred even hardly at all from project-related without the projects? The CCCE audit introduction of animal traction and obviously entertains doubts: cotton. Intensification of traditional agriculture and livestock "...exceeding the production raising proceeds as population objectives appears more as an effect density rises. This basic strategic the project has incurred or as fact is borne out by project- taking place outside of its action implementation experience. While the than as a response to a willful and farming and livestock techniques organized initiative."'/ being promoted by the projects did not cause these problems, neither did The IDA audit shares these doubts. they resolve these problems. Most It is ultimately impossible to project-promoted techniques to unravel cause and effect. Therefore, integrate livestock raising and we, like the CCCE, give Boucle du farming, to settle pastoralists, and Mouhoun and Haute Bassins, as well as to persuade farmers to change their Bougouriba II, the benefit of the use of manure, plant waste, and to doubt. The Audit is not convinced grow legumes and grasses for fodder that, overall, Koudougou Pilot was proved to be impractical for farmers succeesful. and cattlemen. They were promoted without exploring their Sustainability attractiveness to farmers/cattlemen, and without understanding the 16. In view of their resulting conditions under which their adoption from extensification and of the would be attractive. technical/ecological pessimists' critique, are the benefits of the 18. Since intensification four projects sustainable? The Audit proceeds where population density is believes that they are, within bound high, IDA's trzegy of shying away - xvii - from agricultural development in the 20. The Audit agrees that the densely-settled but resource-poor National System was several steps Central Plateau seems unfortunate. forward in adapting classical T&V to Abandoning the negotiated Koudougou BurkLnabe conditions. The Audit project abandoned a chance to finds that further adaptations would intervene where intensification and be desirable to make extension more farm-livestock integration were most responsive to its beneficiaries, and needed and most likely to occur. The to increase flexibility so that, Koudougou Pilot project that replaced where there is little content to be it never became a coherent approach extended, the formal structure can be to the problem in spite of changed and unwarranted public costs intelligent and innovative can be avoided. The Audit does not, supervision. however, share the conviction of the more radical critics that current 19. The fight between extension messages are worthless or technological and ecological environmentally damaging. The Audit optimists and pessimists is reflected believes that, when the historical in disagreements over the best process of population increase and extension approach. During the intensification is understood, places earlier phase and, increasingly exist where there are techniques explicitly, during most of the middle worth extending with current phase under audit here, IDA was extension structures. Hopefully, the promoting the training & visit (T&V) Borrower will continue to adapt these system. Without any doubt, this structures to country conditions. initially made a great improvement by This belief certainly does not introducing order, norms and detract from the priority of problem- accountability. It became oriented, client-sensitive research increasingly clear that T&V's to address the manifest remaining recurrent costs were unbearably high problems, such as how to make and that its elitist contact-farmer livestock and farming intensificatfon system was inappropriate to Burkinabe both sustainable and cheap enough to rural society which exhibits a high interest farmers and herders. degree of solidarity. The Borrower took the appropriate next step, 21. The CRPA staff the Audit adapting T&V to local conditions in worked with unanimously favored area- the National System, which IDA (but based agriculture projects, in which not other cofinanciers) has accepted they have a vested interest, over in the present phase of projects. country-wide ones. In fact, there Other development agencies, notably have been start-up problems the CCCE, argue that the National associated with the country-wide System reforms are an inadequate projects. Assuming that those can be adaptation because they retain three mastered, the Audit believes that faults that T&V has under Burkinabe country-wide projects can be suitable conditions: it is rigidly top-down; approaches to development of rural- it emphasizes structure over content; support activities such as extension, and the appropriate content of research, and possibly credit and extension messages is not there. It others. However, farming and animal is on this final point that the CCCE husbandry are extremely site-specific critique of the National System activities, less adapted than any manifests its technological/ other to centralized planning and ecological pessimism, control. Unless the country-wide programs are decentralized and - xviii - adapted to ensure that interventions required to formulate a sound in technology and services are alternative project was lacking. appropriate for conditions in specific areas, they will not be 23. With the removal of credit from cost-effective. One feature of area- the CRPAs' competence, lack of access based projects is that they have an to long-term credit is a constraint inherent capacity to address this on progress for many projects,-area issue by involving local populations, farmers. The latter is an but usually these projects are very unfortunate cost of adopting sound costly. Occasionally, however, financial policies to ensure circumstances may justify this type sustainability of the agricultural of development. credit system, and means should be developed to overcome critical 22. The Audit also finds that the funding constraints which cannot be decision to change the Roudougou handled without undue risk by the project to a pilot project was regular financial system. hastily conceived, and the analysis 2/ Called Volta Noire at the time of credit approval and until 1983. 2 Jean BENHAMOU. Evaluation retrospective du Proiet de Dvelonpement Agricole des ORD de l'Ouest Volta - Phase II. Paris: CCCE, avril 1990. OED #6878. Cotton develovment nrorams in urkina Paso Cte dIvore. and Togo. June 1987. f CCCE. Op.cit. p. eu PROJECT PERFORMANCE AUDIT REPORT SBCOND 80HGOURIBAIAGRICULTURAL DEVELOPMENT PROJECT (CR. 1097-BA) BOUCLE D HOUHOUN AGRICULTURAL DEVELOPMENT PROJECT (CR. 1284-BUR)!' HAUTS BASSINS AGRICULTURAL DEVELOPMENT PROJECT (CR. 1285-BUR) K0UDOUGOU AGRICULTURAL PILOT PROJECT (CR. 1293-BUR) I. INTR0PUCTION 1.01 This audit of four agricultural development projects -- Bougouriba II (Cr. 1097), Boucle du Mouhoun (formerly Volta Noire)(Cr. 1284), Haute Bassins (Cr. 1285), and Koudougou Pilot (Cr. 1293) -- will attempt to place them, and to evaluate them, in their broader historical context. They may be thought of as the middle of three phases in IDA agricultural lending to Burkina Paso. They followed c first phase which was more-explicitly cotton-oriented. These four projects -- phase 2 -- attempted to broaden the focus from cotton to agricultural development as a whole, especially to grain production, and to the desirability of integrating farming and livestock-rearing in order to maintain soil fertility. In the event, however, these projects realized the bulk of their economic benefits from increases in cotton production.2/ Nevertheless, the intent was important. As the Audit will argue, its results were significant. Not the least of these results was one subsequent shift in rural-development strategy by IDA and a different shift by France, the most important foreign financier of Burkinab6 rural development. 1.02 The four projects are generally considered successful. IDA's appraisal of the phase-three Agricultural Support Services Project (PRESAP) reflects this consensus when it argues, "In contrast to experience in many other countries, the area- specific projects in Burkina Paso have been quite successful."/ If this is so, and the Audit will argue that it is, then why shift from these area-specific agricultural development projects of phase 2 to a phase 3 of risky, national, policy-based projects: the Agricultural Research Project (Cr. 1896) and the Above-mentioned PRESAP (Cr. 1975), both 1/ Called Volta Noire ADP at the time of credit approval and until 1983. 2/ A similar evolution, with similar results, was occurring simultaneously or a little earlier just to the west of the four-project area in Mali. See OED's audit of Mali Sud in Project Performance Audit Report. Mali. Livestock ProJect (Cr. 538-MLI). Mali Sud Agricultural Development Project (Cr. 669-MLI). Mopti Rice II Project (Cr. 753-MLI). ODIPAC Technical Assistance Proiect (Cr. 1174-MLI). Second Mali Sud Rural Development Proiect (Cr. 1415-MLI). November 9, 1990. I/ Staff Appraisal Report. Burkina Faso. Agricultural Support Services Project, 16 December 1988, p. 7, Para. 1.23. -2- appraised in 1988, and the National Village Land Management±4 Project, jointly appraised with the Calsse Centrale do Coopdc6tion Economique (CCCE) and the Gesellshaft fUr Technische Zusammenarbeit (GTZ) in June-July 1990 and still being processed? Does this shift represent fixing something that isn't broken? 1.03 The Audit believes that it can properly assess the success or un- success of the four projects only by answering that question. And doing so requires a bit more historical background. II. MgyLQLiM 2.01 Burkina Paso is at the headwaters of the Mouhoun (ex-Volta) River, as well as of the Combe and of several tributaries of the Niger. It is, however, neither high nor hilly, but a rather flat plain. Soils, like most of West Africa, are predominantly alfasols containing ferrous and aluminous oxides. In a "state of nature," vegetation and decomposing leaf mulch covers the soil, providing organic matter from plant waste and keeping the soil temperature down, which allows an optimal rate of decomposition of the humus, and retarding leaching and erosion. Undisturbed by agriculture, Burkina Faso would have a pretty full canopy of wooded savanna, merging into forest in the far south, where annual rainfall reaches 1200mm., but much less full in the far north, where rainfall is less than 600mm. 2.02 The "state of nature" is a myth, but it helps explain the current agropastoral dilemma. Like all humankind, the ancestors of today's Burkinab6 were once hunter-gatherers, who shifted gradually to extensive bush-fallow agriculture and to transhumant livestock-rearing as increasing population density made these modes of production appropriate. Oral and other history, and the important remnants of gathering that persist, indicate that this shift is recent. Indigenous agriculture was based on three starchy staples: millet in the north, sorghum in the middle, and yams in the south; and on two legumes: ni6b4 (Dolichos unguiculata) and vouandzou or Bambara groundnut. Even today, after a century of government effo rts to promote exotic plants -- mainly cotton, groundnut, maize, cassava and Asian species of rice -- these native plants account for roughly 80% of the areas Burkinabg farm and of the tonnage they harvest. 2.03 Traditionally, farmers cleared the wooded savanna, sparing useful trees, burned the residue, and tilled by hoe to grow these annual crops. Soil exposure to leaching rains and warming sun, as well as the agriculture itself, depleted soil nutrients and organic matter. After 3 to 5 years of cultivation, fields were bush-fallowed for 15 to 20 years, during which time the level of nutrients and organic matter recovered. Agriculture mined soil fertility short-term, but not long-term. The system was sustainable. 2.04 What was traditionally the case is still the case in much of Burkina Faso today. Bush-fallow agriculture, accompanied by transhumant cattle-raising where cattle move to dry-season pastures, is still A/ Gestion de 1'Espace Rurale or Gestion des Terroirs Villageoises. - 3 - sustainable in large areas where population density is 20/km2 or less -- perhaps even 30/km2 or less. Almost all of zhe four-projects' area falls into this category. However, for the Central Plateau, with population density of 60-90/km2, it is not sustainable. In those areas, fallowe are too short to permit soil recovery. Soils have lower nutrient levels and le, humus. They have poorer structure and are more acid, with lower catior, exchange capacity, so that fertilizers get lower responses. Farm yields are lower. There is less vegetative cover. Where people try to farm them, the cycle of degradation is reinforced; when they are abandoned, natural regeneration is slow.5/ 2.05 To understand the changes accompanying the four projects under audit, it is important to understand the historical processes that made these two areas -- the low-density one and the high-dansity one -- different, and how those processes continue today. Why is the Central or Mossi Plateau so much more densely peopled than the rest of Burkina Faso? The difference stems from differing political and social organization. The Mossi who people the high-density area and the other Burkinab6 who peiple the rest of the country are no doubt not so ethnically different as they believe but rather the bearers of different traditions of social and political u:ganization. From the 13th Century, when the first of the three Mossi states was formed by merger of local populations with Fon or Yoruba immigrants from the co3ast, the Mossi formed large socio-political units and assimilated their neighbors or repulsed those who refused assimilation. Their society fostered higher population density than that of their neighbors, and was expansionary.T Consequently, in Burkina Faso, it was on the Central Plateau, home of the three Mossi kingdoms, that human population first reached densities at which a bush-fallow mode of production was not sustainable. From the beginning of the 20th Century, and even before, this led to the "migratory phenomenon," Mossi migrating from their plateau to work in Ghana and in the Ivory Coast. And, more recently, to settle in the adjacent areas of Burkina Faso -- which happen to be the four-projects' area, 1/ This process has been understood for some time. See, for instance, P.H. Nye & D.J. Greenland. The Soil Under Shifting Cultivation. Farnham Royals Commonwealth Agricultural Bureaux, 1960, not specifically about Burkina Faso, but 'y two professors from "Lower Volta". For a devastating analysis of soil degradation under somewhat wetter conditions in the Swango Kwilu region of Zaire, see Louise 0. Fresco. Cassava in Shifting Cultivation. A Systems Approach to Agricultural Technology development in Africa. Amsterdam: Royal Tropical Institute, 1986. g/ See Elliott Skinner. "The Mossi and Traditional Sudanese History," Journal of 77egro History (April 1958) XLIII, #2, pp. 121-31; and "An Analysis of the Political Organization of the Mossi People," Transactions of the New York Academy of Sciences (June 1957) II, #19/8, pp. 740-50; S.A.Balima. Notes eur 1'Organisation de l'Empire Mossi. Paris: M6moires de 1'IHEOM, n.d.; Peter B. Hammond. "Economic Change and Mossi Acculturation," in W.R. Bascomb & M.J. Herskovit (eds.) Continuity and Change in African Cu,7tures. Chicago: Chicago University Press, 1959, pp. 238-56; and Domin:Lque Zahan. "Pour une Histoire des Mossi du Yatenga," L'Bome (Mai-AoGt 1961) 1, #2, pp. 5-22. 2.06 The result of this higher population density on the Central Plateau has been lower soil humus and nutrients, resulting in more land too poor to farm and lower yields on what is farmed, as well as fewer trees and shrubs, resulting in a firewood shortage. The CCCE amdit of Haute Bassins and Boucle du Mouhoun/ believes that this environmental degradation results from a Mossi attitude towards nature that is different from that of other Burkinab6.!/ The Mossi allegedly want to penetrate and exploit like a warrior and the non-Moasi to aurture and sustain like a farmer. This Audit believes that any such attitudinal difference towards nature and land stems from the objective condition of land scarcity and higher population density -- in other words, that economics a ad geography explain anthropology, and not vice versa. 2.07 During the implementation period of the four projects -- 1981 to 1989 -- the most dynamic factor in rural change was undoubtedly in- migration of farmers from the Central Plateau, settling in the four- projects' area to take advantage of better land. Throughout the West African savanna, villages traditionally welcome settlers; their chiefs find settlers land to farm. Growth of a village brings prestige. But the in- migration of the 1970s and 1980s has been so rapid -- sometimes amounting to over 5%/year, that the hosts can no longer assimilate the guests. To the traditional migration-stimulus of soil degradation and over-population on the Central Plateau were added the droughts that began in the 1970s and, more recently, the economic collapse of Cote d'Ivoire. 2.08 The four-projects' area that is now receiving this deluge of in- migrants is the richest part of Burkina Faso. Economically the closest to outside markets as a result of the rail line to Abidjan, western Burkina Faso was where the French first promoted groundnut cultivation.!/ The colonial administration always considered Burkina Paso to be a country without resources and used it as a pool of workers for Ivory Coast and Mali. Development efforts during the colonial period were modest indeed. Burkina Faso's most important foreign-exchange earner has always been workers' remittances and, more recently, foreign aid, but most of its exports came and come from the four-projects' area. Cotton is the most impor-%nt. Bobo-Dioulasso, center of Haute Bassins, has been the center from hich cotton-development spread. From 1952, CFDT (French Textile Development Company) the predecessor of SOPITEX (Fiber & Textile Development Company) was active in Bobo-Dioulasso.1L/ From 1956, 7/ Jean Benhamou. Evaluation retrospective du Proet de Develon,ement Agricole des ORD de l'Ouest Volta - Phase II. Paris: CCCE, avril 1990, 61 pp. + annexes. 8/ Ibid. p. 51 citing M. Benoit. Les oiseaux du 1il. Paris: ORSTOM, 1982, a portion of which is reproduced as Annex 35. j/ See Robert Delavignette. Lee Paysans Noire. Paris: Stock, 1946, 262. pp. JO/ Initially, Bobe was only the administrative center and cotton growing and ginning were further north around Tougan, Kangoussi and Kaya. However, as these proved to be too dry, the operational center shifted to Bobo too. - 5 - government promoted SMDR (Cooperative Rural Development Societies) in each canton, built food stores, promoted plowing with animal traction, hence straight-line sowing and keeping draught animals; crop rotation of cotton followed by grain, hence use of insecticides and chemical fertilizers as well as immense quantities of manure and compost. These are the farm practices worked out by Pierre Viguier at M'Pesoba Farm School just across the border in Mali in the 1930s and 1940s.11/ And, with very few exceptions indeed, they are the extension themes on which the four projects being audited are based. 2.09 The four-projects' area was the center of public rural-development efforts just after independence, but, to keep things in perspective, those efforts, like those of the colonial period, were modest. Anyone interested in the details of how modest they wore may refer to the preparation report for the West Volta Cotton Project (Cr. 225).121 When IDA joined CCCE and FAC in that project in 1971, cotton was clearly the focus, and the CFDT was the implementing agency. It was largely a matter of IDA's joining the French Government in promoting the time-honored vision of agricultural development in the region. 2.10 CFDT always argued that promoting cotton is promotinb food crops because the food crops that follow cotton benefit from fertilizer and from improved land preparation. Nevertheless, worries developed about over- reliance on cotton. Therefore, the later phase one projects -- the 1975-80 Bougouriba Agricultural Development Project (Cr. 496), approved in May 1974; and the 1977-83 West Volta Agricultural Development Project (Crs. 706 & 049, approved in May 1977, cofinanced with CCCE, FAC and Swiss Cooperation -- while still largely cotton projects, had changed names and represented a shift in influence that intensified in phase 2. 2.11 These projects were innovative in other ways as well. Extension was much improved with systematic attention to techniques, and to organizing extensionists' work and checking up on them. IDA can take credit for these il rovements. Moreover, the projects started to get villages to organize village groups (GV), which took over the Job of marketing cotton & agricultural inputs and of managing credit through self- managed markets (march. a auto-g6r&s). To do these jobs, formerly done by SOFITEX, villagers had to be literate and able to keep simple accounts. To this end, the projects instituted functional literacy programs in Dioula. III. PROJECT DESIGN 3.01 All -f the above provided the background for the four projects under audit -- the phase 2 projects. As the respective PCRs indicate, three of them: Haute Bassins, Boucle du Mouhoun, and Bougouriba II, were JI/ See "Centre de Colonisation on Terre Seche de M'P6soba (Soudan)," L'Aaronomie Trovicale (1952), VII, #2, p. 182. 12/ SEDES. L'Etude de Pr6paration du Prolet Coton Ouest Volta (71/2 - ZZL8. Paris, 1969. - 6 - long on continuity with their predecessors,13/ which is not surprising, given the success of their predecessors. The Audit has little to add to the accounts in the respective PCRe, except to note the two nuances of difference between phase 1 and phase 2. 3.02 Firstly, design of these projects took place during or shortly after the drought of the early 1970s that resulted in three successive crop fa4lures, considerable deprivation, and heavy dependence on food aid. It is not surprising that their design embodies a shift in emphasis from cotton to food crops. 3.03 Secondly, the projects reflect growing worries that the agronomic formulae being pushed might not be sustAinable, that they might be leading to long-term loss of soil fertility. These were the years when one heard frequently, and sometimes from serious sources, that the Sahara Desert was marching south at a rate of so many kilometers per year. Bush-fallow sustainability worries that had concerned specialists for decades were also taken up by generalists and self-dusignated ecologists who did not understand them. Consequently, these projects contain more than their share of attempts to get livestockmen to stay put, to integrate ranching and farming, to get farmers to compost manure and other wastes and to put huge quantities on their fields (5 to 6T/ha were usually recommended). To be sure, these counsels were not new. They were part of the extension message of the phase-one projects and even of the bit of extension that took place before independence. Still, reading the prvject documents for all four projects, one cannot help being struck by the sincerity of these sustainability concerns. 3.04 As implementation would show, sincerity was not enough. There were technical design lacunae. With nothing for their cattle to eat and drink in the dry season, how were transhumant herders supposed to stay put? Where were farmers supposed to get the immense quantities of compost, far in excess of what was available? And had anyone figured the cost, in time and sweat, of making that compost, carrying it to fields, and spreading it? They had not, and that is why Burkinab6 refer to West Volta Agricultural Development Project as Cotton II and to Boucle du Mouhoun & HP,tts Bassins as Cotton III. 3.05 Koudougou Pilot is different from the other three, but it started out the same. It was designed, appraised and even negotiated as a regular phase-two project between October 1977 and November 1980. However, its project area was partly on the Central Plateau, which IDA had regarded as 13/ Boucle du Mouhoun was called Volta Noire (Black Volta) when it was approved. The name of that river was changed to Mouhoun in 1983, when Upper Volta became Burkina Faso. Both Boucle du Mouhoun and Haute Bassins are the successors to West Volta Agricultural Development Project (Cre. 706 & 049). They cover the same area, which was split administratively. For the French cofinanciers, the project continued to be one -- The West Volta ORD Agricultural Development Project - Phase II. This is so despite the fact that the things and portions of things financed by CCCE and FAC differed considerably between the two ORD that made up their project. -7- too resource-poor for production-oriented agricultural projects. Koudougou was an attempt to adapt to the Central Plateau the rural-development techniques which IDA-supported projects had been using in the resource- richer parts of the country -- where the predecessors of the other three projects under audit were located. 3.06 Files indicate that the principal worries from identification through negotiations were the same as those of the other three projects. The main one was the high cost of the project. On 5 December 1980, Project.' Department sought to "dispel the doubt that the service we are developing at Koudougou is expensive and unnecessary" because the project would "improve <the extension service's> efficiency through training and the introduction of a T & V system." The worry was evidently not just whether the project was "gold plated," but whether Burkina Paso could bear the recurrent costs of T & V (as well as of fertilizer subsidies, because these had not yet been abolished) and about the rather large (for savanna west Africa in those days) total cost. At green-cover stage, cost was estimated at US$30.2 million, of which $20 million was to be financed by IDA, $2.4 million by GTZ and $7.3 million by another then not identified external financier. Not mentioned in minutes of negotiation are existing, uncovered Koudougou ORD debts.14/ 3.07 Almost as soon as negotiations were finished and the appraisal mission leader transferred to another region, Koudougou RD began to encounter problems, evidently connected with the on-going work on the Agricultural Issues Study, a 2-volume piece of IDA sector work which appeared in yellow-cover on 26 January 1981.151 On 5 December 1980, the Chief of the division responsible acknowledged to the authors of the Study and their Chief, "The real uncertainty...which no one, but no one, once seriously questioned in the review of this project is the constraint of the present social systems on the use of land and how this will affect the degree of adoption of our technical packages." That the project should have been so attacked is surprising since the Agrar- und Hydrotechnik's preparation report16/ was sociologically explicit and complete. Moreover, the appraisal mission included the anthropologist who is the best-known authority on the Mossi of Yatenga (though the mission leader did not find 14/ These are not mentioned in the 21 August 1981 green-cover SAR either and may not have been known at that time. They are, however, mentioned in an internal IDA telex of 23 January 1981, by which time their elimination had become a condition of effectiveness. ORDs (Office Regional de D6veloppement) were the public institutions that unified almost all government rural-development efforts at the district level. The first ORDs were created in the early 1960s, but their coverage of the country expanded slowly. They were summarily abolished in 1986, operated in limbo for a year, and were reconstituted in 1987 as CRPAs (Centres R6gionaux de Production Agro-pastorale), but without, officially at least, their input-distribution and credit roles. 15/ Finally issued on 29 October 1982 as Report No. 3296-UV. 16/ Preparation des deux projfts Koudougou et Bougouriba. Essent mars 1979. - 8 - his contribution useful). IDA cannot have been surprised by "sociological" constraints of which it was initially ignorant. Nevertheless, IDA's resident mission in Ouagadougou also evidently raised this "need for sociological input."17 Otherwise, during late January and very early February there were no serious signs of trouble. GTZ was still a presumed cofinancier and negotiations with the Dutch Ministerie voor Ontwikkelingesamenwarking seemed to promise that the rest of the financing gap would be filled, and that a Dutch grant would cover the Borrower's incremental fertilizer subsidy costs. 3.08 On 11 February, IDA's Acting Assistant Director of West Africa Projects asked his Vice President for "another day to check the work done," having evidently asked for reconsideration of Koudougou in a memo not in files. The Koudougou SAR was assuming that use of nitrogen and phosphate fertilizers on sorghum and millet, with associated husbandry improvements, would raise outputs by 11 to 13 units per unit of fertilizer nutrient. The Agricultural Issues Study was suggesting that the ratio should be 7:1 or perhaps 5:1. A fierce internal discussion ensued. It involved a) the most likely response of sorghum and millet to nitrogen and phosphate applications under project conditions, b) the appropriate value of farm labor for economic-analysis purposes, c) the appropriate economic value of foreign exchange, and d) whether incremental domestic food production should be given a premium value over world-market equivalents because of legitimate concerns over food security. On 24 February, the Loan Officer was still "hopeful with Koudougou." On 12 March, he telexed the Ouagadougou office that "Koudougou is still in intensive care with high probability that its life support systems will be withdrawn." A big discussion was held on 17 March. On 20 March, a memo announced that, given revised assumptions, the ERR was projected to be an unacceptable 7-82, not the 242 projected in the SAR, so that IDA could not go ahead with the project. The Resident Mission was informed the same day. The Borrower was informed officially in 1 April, the two prospective financiers on 3 and 8 April. Germany formally withdrew from discussions on 8 July. 3.09 This is a rare case of a project, not just in the IDA lending program but already negotiated, being thrown out on technical grounds. How difficult such a st-p is for IDA is only partly evidenced by the flurry of explanatory memos in files. What were the grounds? 3.10 It was information about sorghum and millet responses to fertilizer from the Agricultural Issues Study that undermined confidence in the SAR's projected response rates, which caused IDA to change its mind on those projections, which, in the absence of compensating adjustments, made the projection of economic returns unacceptably low. This information seems to come from three sources: the FAO Fertilizer Programme's annual report on (then) Upper Volta covering 1977, 78 & 79; Peter Matlon's paper, subsequently delivered at an Abidjan meeting, on ICRISAT findings about farmers' field tests of fertilizers and improved practices; and estimates in the SAR of the Bougouriba II project.18/ The Study authors, in 11/ See letter from headquarters to Resident Representative of 22 December 1980 defending the appraisal mission. IS/ Agricultural Issues Study. p. 79, Para. 1.5.22. See entire fertilizer section (pp. 73-96). - 9 - interpreting these data from numerous sites in Burkina Faso, recognize that fertilizer responses are site specific and husbandry specific. They divide the response data into "center" and "southwest" and then generalize high and low responses per crop and per these two regions. Their generalizations were informed by observations from IRCT and IRAT about unsustainability of savanna annual cropping and decreasing long-run responses to fertilizers.191 IDA chose to believe the lower fertilizer responses consistent with the Study rather than the higher ones proposed by the preparation report and the SAR. The northern part of the Koudougou project area accounted for 48% of the area and 862 of the farm families. It was considered too dry for cotton, and the project's agronomic package was based on increasing sorghum and millet production. Fertilizer response was critical. This reinforced IDA's traditional view that there was no attractive agronomic package for the Central Plateau.20/ 3.11 All of this had little to do with Koudougou's southern half, where population density was 9/km2 as vs. 61 in the north. Koudougou south is similar in resource endowment to the Bougouriba project area and to the richer parts of Hauts Bassins and Boucle du Mouhouns sociologically, it is similar to Bougouriba. Moreover, the Secretary General of the Ministry of Agriculture specifically requested that IDA continue a full-scale project in Koudougou south. IDA did not, however. The Issues Paper of 6 November 1981 explains, that the decision was based on "...uncertainty as to the willingness of the different types of farmers (Gourounsi, Mossi, Peul) to accept improved agricultural practices.... Doubts about the acceptance rate find their origin in the low level of development (even in a relative sense) of the area's original population (Gourounsi) and the looser structure of their society. This situation was explained to the mission by the Voltaic authorities and the mission found it clearly in evidence during its field trips. Moreover, during the first Bougouriba project we have had problems with comparable population groups (the Lobi). Apart from the original population, there are Mossi immigrants in the area. Mossi people are thought to be more receptive to improves agricultural techniques. However, Mossi migrants are usually young and poor and often lack the means to acquire inputs and equipment even if they were willing to do so. Moreover, they consider their stay in the south temporary and ... may therefore have little incentive to make investments in labor and money."21/ 19/ IRCT. "Comit6 de Recherche Agronomique." Bobo-Dioulasso, 1980 and IRAT. Rapport de Synthbee 1979. March 1980, discussed at Agricultural Issues §tudy, p. 81, Para. 1.5.24 20/ This was the point of putting productive agricultural projects in the west, while the Agricultural Development Fund projects (Cre. 317, 640 & 1218, approved June 1972, June 1976 & March 1982 respectively) provided poverty-alleviating improvements (drinking wells, village food storehouses, "bas fonds" development) for the Central Plateau. 21/ "Upper Volta. Appraisal of the Koudougou Pilot Project. Issues Paper," - 10 - IDA had clearly made the implicit decision not to go ahead with a full- scale project in the south well before the appralial mission. In view of the future Borrower's known preference for a full-scale project in the south, the 3 September Project Brief says (p. 7) "the Government may not be fully aware that the project will be considerably reduced in the process." What the September-October 1981 .e-appraisal mission proposed for the south was an applied research and extension intervention covering only 0.33 million ha, with a Dutch veterinary intervention over the rest; this was also a much smaller IDA intervention than the Dutch had anticipated. 3.12 In the event, Koudougou was down-graded from a US$30 million project to a US$9.9 million pilot project. Instead of seeking agricultural, indeed rural development over the entire project area as the other three projects under audit did, it was cut down geographically to two very small sub-divisions,22/ one in the densely-settled north and one in the sparsely-settled south. It was cut down substantively to an applied- research project, looking for answers to the agronomic problems in the north and, in the south, both looking for answers and looking for an extension approach that would persuade its people with their "low level of development" to change. 3.13 Although the pilot, like the other three projects under audit spoke of integrating agriculture and livestock-raising, it separated them by donor. The Dutch were to finance and provide technical assistance for livestock, as well as M&E; IDA was to finance agriculture. That financing was to be parallel. As our discussion of Implementation (below) will show, this design feature was not conduci. to integration. Implementation was parallel too. The Dutch cofinancing partners were party to the change, but complained to the Audit that consultation was inadequate. Whatever may have appeared to be the case at the time, results have shown that the Dutch partners were not fully convinced about the new and radically different design of the project. As implementation will show, the parallelism was a distant one. 3.14 The reformulation decision, based on IDA's pessimism regarding agronomy in the north and "sociology" in the south, is an impressive contrast to the usual excessive optimism during the project-review process. In this Audit's opinion, the final cause of reducing output projections, and consequently to down-grading Koudougou to a pilot project, was not so much some specific research result as a pessimistic assessment of savanna- zone agriculture based on wider Bank experience. If the warp of this tissue of experience was worry about "desertification" of the savanna, the woof was a growing conviction that integrated rural development projects do not work. 3.15 Regarding "desertification," there was the evidence from the northern part of the Koudougou project area, the Yako sector, ecologically part of the Central Plateau, with high population density, its soils' fertility reduced by prolonged cultivation without enough fallow so that yields were low, its woody vegetative cover reduced so that firewood was getting scarce. With the evidence of repeated droughts and concern that 22/ "Cantons" in the colonial system or "kafo-oul in the pre-colonial system. - 11 - reduced rainfall might be a permanent phenomenon, IDA staff's assessment of agricultural potential turned radically pessimistic. Moreover, they seem to have 'eared that the same thing might happen to the sparsely-settled and relatively rich southern part of the Koudougou project area. 3.16 Regarding the workability of integrated rural development projects, by 1981, IDA staff (and the rural development community in general) had had enough experience to appreciate the management difficulties.23/ These projects had been the panacea of the 1970s. By the 1980s, it was becoming clearer that it was hard to manage the many components, hard to get the line agencies to transfer their best technicians to autonomous project agencies, and hard to get the cooperation of those agencies. In 1981, IDA was still promoting such projects in aear- by Nigeria, but that promotion stopped not long thereafter. In Burkina Faso, the critical re-assessment seems to have occurred in 1981. 3.17 The down-grading of Koudougou was the first effective result of the worry about integrated development projects as an organizational form, and of worry about the content of the savanna rural projects because of their effect on soil fertility. Was this design change wise? The Audit will have something to say on the subject after having loc;ked at implementation. IV. IMPLEMENTATION 4.01 These four projects seem to have exceeded their projected production increases (see Impact below) but without spending the anticipated amount of money.41 Of the US$34.5 million of IDA credits approved for the four projects, 38% was cancelled. This amounts to 23.3% for Bougouriba II, 29.2% for Koudougou, and, for the twin projects that were successors to West Volta ADP, 61.7Z for Haute Bassins and 62.5% for Boucle du Mouhoun. When the Audit visited these projects, all claimed to be strapped for funds and equipment. Yet all credits had been closed less than two years earlier, cancelling US$13.7 million in unspent IDA balances. Needless to say, losing such a cheap source of foreign financing is undesirable unless the purposes for which they were to have been spent were really socially worthless. Why did this happen? With the few exceptions noted below, the PCRs shed little light on this subject. 4.02 All of the projects bridged the period of the Sankarist revolution (August 1983-October 1987). One of the many effects of that revolution was 23/ The history of this growing consensus is covered in IBRD.OED. Rural Developments World Bank Experience. 1965-1986. Washington: 1988. This document synthesized the consensus and, from the time it began to be researched and written in 1984/5, contributed to the development of the consensus. 24/ Before we jump to the conclusion that the projects used public monies even more efficiently than expected, remember that most of the non- farm-production goals, e*g* integration of livestock and farming, were not significantly achieved. See Impact below. - 12 - to strike fear into the hearts of civil-servant managers, like the Directors cf the four projects. Such officials were unaccustomed to being held publicly responsible for their stewardship, particularly to being haled before People's Revolutionary Tribunals. The Koudougou PCR25/ notes the chillingly effect of these tribunals on Director initiative.267 The other PCRs cov.ld have mentioned it too. 4.03 Another part of the revolutionary impact was the abolition of the project-implementing organizations, the ORDs, in 1986. The ORDs did not disappear when they were abolished. They went on functioning; project employees went on getting paid. However, there was very great uncertainty. Worst of all, the situation was in limbo for more than a year. Finally, in 1987, the government created CRPAs to take over most of the functions of the ORDs. However, the CRPAs were not to handle either distribution of agricultural inputs -- previously a major source of income to the ORDs -- or credit. 4.04 In fact, the CRPAs continue to handle both credit and input distribution to some extent. They must do so because they clone have agents in villages. The CNCA, to whom the farm-input credits are owed by the farmers through their GVs, cannot collect them. The CNCA has only 6 offices in the four-project area.27/ It must, perforce, work through CRPA workers to collect outstandings. Moreover, since CNCA does not have the articulated rural network which the CRPAs do, its capacity to make new loans is extremely limited. For all intents and purposes, most villagers ;L/ ES, p. iii & p. 9, Para. 8.01 26/ Perhaps the Sankarist attempt to make project officials publicly accountable had justification. An official at Boucle du Mouhoun told the Audit that, at the end of the IDA loan, when the foreign M&E officer left and the M&E system was refocuseed to give information pertinent to project management "& temps utile", lots of unexplained consumption of gasoline & spare parts, and excessive missions were discovered. 21/ Caisse National de Crddit Agricole. Three in Boucle du Mouhoun (D6dougou, Solenzo, Boromo); three in Haute Bassins (Bobo Dioulasso, Hound&, Orodara); none in Bougouriba II or Koudougou which are "served" by CNCA out of its Ouagadougou central office. - 13 - who are not members of Caisses Populaires28/ now have no access to formal credit at all. The CRPAs credit units have one residual function only, to collect CNCA's remaining outstandings.29/ SOFITEX, too, has no village- level agents. Therefore, SOFITEX depends on CRPA extension agents to help it estimate the demand for seed, fertilizer, pesticides, ULV sprayers, and other inputs. 4.05 As if the above-mentioned problems were not enough disincentive to CRPA staff, there has been a jurisdictional dispute between ministries. Part of the Sankarist revolution was peasant empowerment, and this led to the creation of the Ministry of the Peasant Question, which has since become the Ministry of Cooperation Action. That ministry has been made responsible for formation and management of village groups (GV), an essential extension function formerly the job of the ORDs. However, the new ministry is essentially political and not actually staffed to do the job, which is still actually done by the CRPA, under the Ministry of Agriculture. In the light of all this confusion, which is still not completely sorted out, it is not surprising that projects' implementation was slow. Rather, it is surprising that the ORD/CRPA staff carried on and that they managed to implement the projects at all. 4.06 Coupled with the institutional turmoil described above, implementation of three of the projects had to contend with disagreements and/or misunderstandings between cofinanciers,30/ vith long delays in procurement and in assignment of foreign experts who were not, in some cases, suitable. 28/ These rural credit unions are part of the worldwide Reiffhiesen movemenc. In Burkina Paso, they are grouped into two unions: the URCPB (Union R6gional des Maisses Populaires du Burkina) and the UCECB (Union des Caisses d'Epargne Cooperatives du Burkina). The differences (URCPB has only farmer members and one bank/village; UCECB accepts merchant and civil-servant members and has one bank for several villages) need not concern us here. The movement has been growing slowly but steadily for over 20 years. The banks are quite solid. URCPB is centered in the Bougouriba project area, where it has 12 banks and 18,500 members (out of a population of ca. 400,000) with savings of ca. CFAF 500 million (US$2 million), of which only 20% is lent to members and 80% is in banking instruments. Repayment periods are excessively short for plows, oxen, etc., discouraging all but short-term borrowing. Interest paid on deposits is low -- 3% or less -- but transaction costs are high for the many, small accounts. UCECB has some branches but much less influence in the Koudougou and Haute Bassins project areas. Z9/ Since they are collecting but cannot respond to requests for new loans, CRPA credit officers told the mission that they resort to a number of lies: you can't get any new loans because some people in the village are still in default, but, when no one is, because some people in neighboring villages are in default. 30/ The PCRs do not explore these matters. - 14 - 4.07 The most visible disagreements were over extension philosophy. The Audit confirmed reports of high-decibel altercations between a CCCE representative and the IDA Special Adviser during a joint visit to Boucle du Mouhoun in October 1985. The Borrower's representatives complained to the Audit about this behavior. The Audit pointed out, however, that this also helped the Borrower realize that how Burkinabd extension should be done has to be a Burkinab4 decision. The disagreements helped lead to a national consultation and agreement on the National System, a significant improvement, adapting T&V to Burkinab6 conditions, which IBRD has agreed to support through PRESAP. (More at Issues and Findings below.) 4.08 Implementation of Koudougou Pilot was also marked by misunderstanding between IDA and Dutch cofinanciers. The Pilot that was concocted to save something when the regular Koudougou project was dropped was not strong on coherence to begin with. It called for the Dutch to look after M&E and livestock, for which intervention was pretty much limited to veterinary medicine. Both the Dutch and the IDA SPN missions early recognized that, if farm-livestock integration was to be attacked in a serious way, the problem was animal feed. A series of "reformulation" missions and long periods of indecision ensued, culminating in what the Dutch called their appraisal mission of late 1986, when the IDA credit had already been effective for three years.31/ At the wrap-up for the September 1986 SPN mission, the Dutch Second Secretary complained that IDA and the Borrower had substantially amended the project without consulting or informing the Dutch. In fact, IDA and the Borrower were modifying Koudougou Pilot and were laying the groundwork for the subsequent shift to a country- wide extension & research strategy in phase 3. Meanwhile, Dutch assistance for M&E was discontinuous. Since pilot projects are supposed to lead to something else, M&E might be thought to be important, but IDA SPN missions complained that the SAR conception of M&E was too complicated. Perhaps it was too complicated to monitor something that was being constantly redefined. Only in 1988 did the Dutch part of the project really get going. By the end of 1988, when IDA was deciding not to extend the project, the Dutch had only disbursed about US$ 0.5 million, two-thirds of that for M&E and one-third for livestock development. (see PCR, p. 16) 4.09 Meanwhile, as the Dutch were trying to settle on a role in livestock and M&E, the Borrower vas implementing the IDA-supported pilot agricultural research project. Actually, most IDA money was going for setting up project headquarters. By the intended completion date, October 1987, according to IDA disbursement records (see PCR, p. 16), just over 1/3 or about US$2.5 million of the IDA credit had been disbursed. Of this, roughly 702 had gone to project headquarters, 182 to construct the northern and southern zone PAPEMs, and 122 to research. The Pilot's deliberate concentration on the two PAPEM sub-divisions caused dissension within the Koudougou ORD, which had just become a CRPA. The staff in the chosen sub- divisions had all the funds they needed; those in the rest of the CRPA had very little. The IDA agronomist task manager worked out a formula for re- extending the project to all of Koudougou CREPA. But on 19 May 1987, IDA suggested to the Agriculture Minister that regional projects would not be 31/ See Dutch Embassy's letter to IDA staff of 24 October 1986 transmitting draft TOR for appraisal mission of Conrad von Toonen. - 15 - renewed. Thenceforth debate was over how long to extend Credit 1293 to permit an orderly transition to support under PRESAP ((Cr. 1979, effective in May 1989). The Borrower requested an extension to September 1989 but IDA only granted one until March, creating a hiatus and further administrative confusion. Nevertheless, the extension permitted another US$2.9 million to be disbursed, about US$1.9 million of it for the extension in space to the rest of Koudougou CRPA. 4.10 IDA's reversion to an area agricultural-development project in 1987 roughly coincides with the real take-off in Dutch livestock and M&E financing. Reports from the Borrower's field staff suggest quite a lot of confusion, reminiscent of the normal problems of project start-up, but in the project's fifth year. The Dutch had manifestly taken a at- : forward from the veterinary focus in the appraisal documents to fc-us on the problem of animal food and water in the dry season, which is clearly the key to farm-livestock integration. In any case, they were distressed by IDA's decision in 1988 not to extend the project beyond March 1989. 4.11 Long procurement delays have already been mentioned as a theme of these projects' implementation. The public rural-development services in Burkina Faso obviously have ample experience with IDA procedures from earlier lending. There is no suggestion that their inexperience was a factor. The Sankarist revolution may have had an effect. What the Audit can be sure about, however, is the effect of what a Boucle du Mouhoun Director called the complexity of the project's financial arrangements."32/ The Staff Appraisal Reports (SAR) for Boucle du Mouhoun and Hauts Bassins, the projects co-financed by IDA, IFAD, CCCE and FAC, show awareness of the difference between IDA and French procurement procedures. The first glance at SARs and legal documents suggests that there was to be joint33/ financing of certain categories of construction, of vehicles & equipment, of salaries & operating expenses, of technical studies, studies, audits & bursaries, and of future sub-projects. However, the joint part was limited to a portion of these categories. Some portions of each (except future sub-projects) are reserved to either IDA-IFAD or CCCE-FAC; other portions were genuinely joint. In each case, the SAR and/or the legal documents provide some description and/or the dollar amount of the reserved sub- categories. The ticklish item turned out to be materials for non-contract construction (most of it), and vehicles & equipment. These were to be financed according to IDA guidelines "and through procedures satisfactory to France for contracts equivalent to US$0.3 million to be financed 100% by France."341 There may have been documents showing what the reserved $0.3 32/ "complexit6 du montage financier du projet" 33/ i.e. between IDA-IFAD and the two French organizations, CCCE & FAC. IFAD accepted IDA procurement procedures. 34/ Volta Noire ADP SAR p. 19, Para. 3.20 and Haute Bassins ADP SAR, p. 14, Para. 3.21. - 16 - million was, but the Audit has not found them. According to the Schedule 1's of the respective legal documents, a small portion of vehicles & equipment -- expected to cost $0.075 million in the case of Boucle du Mouhoun and $0.067 million in Hauts Bassins -- was to be jointly cofinanced 60-40, while $0.34 million and $0.285 million respectively were to be entirely financed by IDA & IFAD. While all of this seems needlessly intricate, it is, well, almost explicit. 4.12 In fact, however, during implementation, procurement arrangements caused great confusion and delay, especially for Boucle du Mouhoun. Project management drew up ICB documents for project vehicles, including motorbikes for extensionists, and sent them to the cofinanciers for comment. Receiving "no clear reaction,"3S/ bids were called for, ..ceived, and evaluated before France refused to finance procurement from the lowest qualified bidders, all Japanese firms. and IDA and the Borrower insisted on it. After some tergiversation, the markets were split, with IDA-IFAD financing purchase from the lowest qualified bidders and CCCE financing purchase from Peugeot. Not only did the process have to be repeated so that getting the vehicles was deleved by years, but the Borrower complains that the French vehicles were less robust than they had specified, particularly in the case of the motorbikes, which they reserve for in-town use, and more expensive than if French financing has been untied. 4.13 How could this confusion have occurred? From the SARs and well as from common sense, it is apparent that the parties were aware of the tied- ness of most French financing. Burkinabd had procured under French financing before. If the Borrower's representatives inexplicably failed to see trouble coming, diligent supervision by the cofinanciers should have seen the problem with the bidding documents. It did not, which caused a good deal of embarrassment and, worse, major delay. 4.14 Misunderstandings with cofinanciers were not limited to procurement. Boucle du Mouhoun management, for instance, tried without success to find out how much a FAC-financed house cost. For Boucle du Houhoun and Haute Bassins, reported project costs are estimates based on what the Borrower thinks FAC spent.31 When trying to assess the impact of these projects, deficiencies in FAC financial reporting to the Borrower and to the cofinancing partners is probably trivial. Project authorities must have a pretty good idea how much the FAC Boucle du Mouhoun house cost, and what total project costs are. Evaluating the impact of Koudougou Pilot when the Dutch part is still going on is another matter. 35/ the Boucls du Mouhoun Finance & Administration Director. 36/ FAC financing was provided for both Boucle du Mouhoun & Hauts Bassins by two agreementst 314/C/DPL/82/VOL of 6 December 1982 for CFAF 200 million, and 239/C/DPL/84/VOL of 4 August 1985 for CFAF 100 million. - 17 - 4.15 This catalogue of implementation difficulties makes the generally favorable results even more remarkable. The Audit wishes to point out that, in addition to the results mentioned under Impact below, several noteworthy things went right during implementation. We will cite the two most important. Firstly, a fertilizer price policy dialogue, conducted largely in conjunction with Bougouriba II supervision, resulted in elimination of fertliser subidies, which had been a very heavy fiscal burden indeed. The success of the dialogue was confirmed by an IDA fertilizer credit (#1550), effective in November 1985. The "cotton" NPK mix subsidies, which had reached 641 of cost in 1981, were progressively reduced and then eliminated in 1988, and urea subsidies were eliminated in 1986. Fertilizer consumption declined 6 to 72 in 1984/5 and 1985/6 as farmers adjusted their use to higher prices, then resumed its upward path.37/ At the margin, farmers were more careful in their fertilizer use and more inclined to use manure. 4.16 Secondly, projects' implementation fostered the first steps in the adaptation of T&V extension to Burkinabi conditions through the National System. A national seminar for reflection on the extension system in 1984/5 led to three years of experimentation with the "test" system (ORTVA) in three CRPAs and now to the country-wide generalization of the system, confirmed by the IDA PRESAP credit (1975) in 1989. This considerable achievement, further discussed under Issues and Findings below, came more from Burkinab6 initiative than from that of IDA, but it was a part of implementation of the four projects under audit nonetheless. These two important policy reforms show that project loans can achieve as much structural adjustment as "policy-based" loans. As discussed under Impact below, they stimulated production of more cotton and some more food crops too. V. LPACT 5.01 The Audit has little to add to the favorable results of these projects reported in the PCRs, confirming earlier favorable findings on Boucle du Mouhoun and Haute Bassins by an OED Impact Evaluation StudyL8/ and the recent ones of CCCE's ex-post audit mentioned above. To review, 27/ For figures and an analysis of fertilizer de-subsidization, see "Projet Engrais Vivriers: Quelques Donn6es Chiffr6es our les Engrais N P K et Ur6e," Ministbre de l'Agriculture et de 1'Elevage. Direction des Intrants et de la Mechanisation Agricole, juin 1990, 12 pp. 38/ OED #6878. Cotton development programs in Burkina Faso. C8te d'Ivoire, and Togo. June 1987. - 18 - Boud* du Mouhoun and Rauts Bassins' implementation periods were accompanied by large increases in cotton production and increases in grain production, generally exceeding projections. Their PCRe carefully not out increases eteming from in-migration which occurred independently of the projects. The remaining increases are enough to give an economic IRR of 202 for Haut* Bassins and of 352 for Boucle du Houhoun.L/ Bougouriba II claims an economic IRR of 182, based principally on increased grain production; there is no netting out of in-migration effects, but then there was probably less in-migration than in other project areas. Koudougou Pilot does not attempt to estimate any output effects but claims that its applied research came up with interesting results. None of the projects claim success in farm-livestock integration. Livestock interventions were effectively limited to veterinary programs. Bougouriba II and Boucle du Mouhoun had some success in p.rsuading farmers to build stone contour barriers. S. 02 Post-PCR production data provided to the Audit by Bougouriba II show that 1989/90, on which much of the returns' recalculation was based, was an exceptionally good year. Cotton production has dropped in Poni, and generally results are more modest than the modest increases claimed by the PCR. Incorporating these corrections would lower the economic IRR modestly but would not change the basin PCR conclusion that the project is a modest success. 5.03 After visiting one of Koudougou Pilot's two applied research centers (PAPEMs), the Audit is not so enthusiastic about the good sense and applicability of work done there as the PCR is. In December 1985, ICRISAT's Principal Economist wrote of the project's PAPEMs to IDA's Resident Representative, "It is questionable whether the PAPEMs, ...given the generally inappropriate orientation of their technology testing, can usefully serve as a means to instruct extension agents in farmer adapted cultural practices." What the Audit saw going on at Koudougou's Kassou PAPEM and at Boucle du Mouhoun's Ki6 PAPEM was largely outside the realm of the possible for farmers outside the fence, which is terribly regrettable becau: i these I-stitutions seemed to promise so much in bridging the gap between searchers and farmers. This also greatly weakens the Koudougou Pilot PCR's contention that the project vas successful, based so much as that judgement is on the value of the PAPEM's applied research. 5.04 The Audit is impressed, as CCCE's audit is, that almost all production increases came from increases in area cultivated -- what CCCE's audit calls extensification. Yields increased little, and increased urea 39/ The CCCE Ouagadougou PCR of June 1989, based on reports of the two CRPAs without netting out the effects of in-migration, arrives at economic IRRS of 36% for Haute Bassins and 55% for Boucle du Mouhoun. - 19 - consumption accounts for most of that increase.O/ If the "quantitative" objectives were often exceeded, what CCCE calls the "qualitative" ones -- farm-livestock integration, extension for herders, literacy, increased use of organic manure -- were not exceeded. For these aspects, there is little to show. 5.05 So the verdict of success, at least for Boucle du Mouhoun, Haute Bassins, and Bougouriba II, is built on what CCCE calls "extensification" -- expansion of the traditional farming system plus increased ox-plowin insertion of cotton into the cycle, and fertilizer and pesticide use4 / (all controlled for the effect of in-migration). To what extent is that phenomenon a secular trend that would have occurred even without the projects? The CCCE audit obviously entertains doubts: "...exceeding the production objectives appears more as an effect the project has incurred <subi> or as taking place outside of its action than as a response to a willful and organized initiative."421 AQ/ This contrasts with earlier OED findings. For the predecessor projects, cotton yields had been rising as farmers increased their use of fertilizer, pesticides and improved seed; other yields, mainly grain, had been stagnating. Markets for these food crops were (and are) erratic, and are made more erratic, from the farmers' point of view, by the intervention of OFRACER (National Cereals' Office). Where spare land was available, farmers expanded but saw no interest in spending cash for fertilizer on a crop they could probably not sell profitably. There was, however, intensification for cotton. In the 1980s, this changed, a fact that was not yet evident to OED's impact evaluation of these and other cotton projects (Cotton Development Programs in Burkina Faso. C8te d'Ivoire and Togo) in 1988. As figures from the Hauts Bassins and Boucle du Mouhoun PCRs marshalled in Annex I show, between 1982/3 and 1987/8, cotton yields were flat, while area continued to grow. Over that period for Boucle du Mouhoun, 1/4 of the increase in cotton production stemmed from yield increassa, 3/4 from area increases. For Hautse Bassins, yields declined and all increases came from larger cotton area farmed. Similar figures are not available from the ither two projects. The major reason for the fall-off in cotton intensification was removal of fertilizew subsidies during this period and the farmers' rational response to that removal. The price of the "cotton" NPK mix in CPAP/kg from the Boule du Mouhoun CRPA over the period in question was: 1982/3 1983/4 1984/5 1985/6 1986/7 1987/7 45 70 90 100 125 105 Ai/ As per footnote above, removal of subsidies after 1982 lead to much more judicious use of these inputs. A2/ CCCE. Op.git. p. v. - 20 - The IDA audit shares these doubts. It is ultimately impossible to unravel cause and effect. Therefore, we, like the CCCE, give the Boucle du Nouhoun, Haute Bassins and Bougouriba II the benefit of the doubt. 5.06 In view of their resulting from "extensification" and of the technical/ecological pessimists' critique, are the benefits of the projects sustainable? The Audit believes that they are, within bounds set by population density. Neither cotton production nor ox-plowing need lead to soil fertility depletion and/or deforestation as some critics maintain. Whether they do or not depends on what else is grown in association or in rotation with cotton, and, ultimately, on the adequacy of the fallow. That is almost entirely a function of population density. 5.07 Those who view cotton and ox-plowing as ecologically destructive (and sometimes as taking food out of the mouths of babies in order to generate filthy foreign exchange) argue that these innovations which are key to the projects under audit do affect fallows when population density is held constant -- that they do so 1) by increasing the amount of land a family can till and 2) by fostering a "cleaner" fa'ming that removes trees that hoe farmers would work around, thus leading to a poorer grass fallow instead of a tree and bush fallow. There is some truth to this argument, but not too much. ICRISAT evidence suggests, "that animal traction ... is associated with only 10-20Z gains in area expansion"43/ or that they are negligible.44/ The expectations of the 1960s that animal traction would double cultivated area per family have just not materialized as other constraints than preparatory tillage emerge. Moreover, as ICRISAT reports and anyone can see and as the PCRs here confirm, adoption of animal traction has not been associated with intensification. Farmers still value n6r6, karit6 and baobab trees and plow around them. So, at worst, the project-promoted innovations have moved the frontier at which the area's bush-fallow systems become environmentally destructive to a somewhat lower human population density. 5.08 There are clearly unresolved problems about farming and soil degradation in the area. Where population density is low and fallows are long enough, they are either short-term problems resolved by fallowing or they are minor. In the lowest population density areas, like the south of Bougouriba and Koudougou and remote parts of Hauts Bassins and even Boucle du Mouhoun, the traditional agricultural system is working well enough that the CRPAs cannot generally persuade people to change at all. This is attributed to the curious "sociology" of the Lobi, Gourounsi, or whoever lives there. It is in somewhat higher population density areas with good market access that the projects have had their major successes. And the projects did not tackle the higher population areas where fallows are too short, like the north of Koudougou. In the Audit's opinion, this is where 43/ ICRISAT Principal Economist P.J. Matlon to IDA Resident Representative of 4 December 1985, Para. 5. A4/ John McIntire, Daniel Boursat & Prabhu Pingali. Crop Livestock Interactions in Sub Saharan Africa. Washington: IBRD, forthcoming in 1991. - 21 - i%. is most important to tackle agricultural problems, and also where intensification has the highest pay-off and is most likely to occur. VI. ISSUES AND FINDINGS 6.01 The Audit will concentrate on five issues, on which it will provide its findings, concluding with a final comment on monitoring and evaluation. A. Ecological effect. 6.02 Did the projects contribute to soil degradation or other negative effects on the ecosystem? Are their benefits sustainable? 6.03 The Audit finds that soil degradation and other negative ecological trends stem largwly from increases in population density and hardly at all from project-related introduction of animal traction and cotton. Intensification of traditional agriculture and livestock-raising proceeds only as population density rises. This basic strategic fact is borne out by project-implementation experience. While the farming and livestock techniques being promoted by the projects did not cause these problems, neither did they resolve these problems. Most project-promoted techniques to integrate livestock raising and farming, the sedentarize pastoralists, and to persuade farmers to change their use of manure, plant waste, and to grow legumes and grasses for fodder proved to be impractical for farmers and cattlemen.!5/ They were promoted without exploring their attractiveness to farmers/cattlemen, and without understanding the conditions under which their adoption would be attractive. 6.04 Since intensification proceeds where population density is high, IDA's strategy of shying away from agricultural development in the densely- settled but resource-poor Central Plateau seems unfortunate. Abandoning the negotiated Koudougou project abandoned a chance to intervene where intensification and farm-livestock integration were most needed and most likely to occur. The Koudougou Pilot project that replaced it never became a coherent approach to the problem in spite of intelligent and innovative supervision. B. The Koudougou downgrade. 6.05 Was it a wise decision? 45/ Interaction between farmers and cattlemen is traditional in savanna west Africa. Under the projects, their predecessors and their successors, many farmers have become owners of bullocks for plowing, but almost all coatinue the traditional interaction of entrusting these animals to Fulani specialists who transhume them to dry-season pastures, rather than the integration of keeping them year-round, growing their dry-season feed, and using their waste products. See McIntire at al. Q2-cit. - 22 - 6.06 The Audit is loath to complain on a rare occasion like this one where IDA staff takes seriously technical questions like sorghum and millet response to fertilizer and sociological receptivity of Gourounsi to change. Nevertheless, complain we will. The decision to abandon the full US$30 million project does not look wise in retrospect. The decision may have been based on misreading or misunderstanding data. 6.07 Regarding fertilizer responses, the SAR was, no doubt, sanguine. However, projectr4 sorghum and millet increases were not just responses to fertilizer but to fertilizer and husbandry improvements and water- harvesting/anti-erosive measures. It was not reasonable to compare them with FAO Fertilizer Program and ICRISAT research results. The critique of the SAR might not have carried the day had not the mission leader been on the point of transfer, the agronomist been on his farewell mission before leaving IDA and not involved after the field work, the back-up agronomist been a consultant unfamiliar with IDA intrigue and not present at Headquarters for most of it, and the Programs' division keen on showing the utility of its Agricultural Issues Study.46/ Such are accidents of bureaucratic life. IDA was trying to be technically serious. 6.08 Current evidence suggests that this technical critique was close to the mark. A. IFAD agroforestery, water and soil conservation project administered by BOAD (West Afri:an Development Bank), was started in Yako sub-division, the northern pilot sub-division of Koudougou Pilot, in 1988. With a technical package virtually identical to the original Koudougou project, it is, according to 8OAD supervision reports, getting responses to fertilizer, husbandry improvements, and water-conservation works consistent with Atricultural Issues Study projections. Early results, however, suggest that these responses are likely to make the investments viable. 6.09 Regarding "sociology," the appraisal mission's expertise was in Yatenga Mossi sociology, not Gourounsi sociology. So IDA's decision to decline the Borrower's request to go ahead with a full project in the sparsely-settled center and south of Koudougou was based on amateur sociology. If it had been consistently applied, it would have forced a similar down-grading of Bougouriba II, approved in January 1981, the very month the great Koudougou downgrade debate was going on. Kordougou south was recognized to have the same sociological problems as Bougouriba (see above, Para. 3.11). A more environmentally-sensitive sociologist would have recognized, as a forthcoming IBRD publication does,7/ that this social behavior stsms from factor prices (especially that of land) which are consequences of low population density and poor market access. 6.10 The rationale for downgrading the Koudougou project, therefore, appears questionable, but of more concern is the make-shift character of 46/ In IDA, such studies would normally have been carried out by the agriculture division, not by the programs' division. In this instance, the agriculture division was too busy with lending operations. The programs' division chief hired an agricultural economist from Stanford Food Research Institute to do the job. A7/ McIntire at al. Op.eit. - 23 - the Pilot that replaced it. Naturally, as a key division chief began,481 "The cancellation of the Koudougou Agricultural Development Project after negotiations vas embarrassing and raised the questions of what vent wrong and how we can minimize the occurrence again." IDA, having made its decision, could not back out cleanly but sought a face-saving formula, the Pilot. Under Implementation above, we have recounted some of the hesitations, the shifts in conception, and the evidences of lack of conviction in Koudougou Pilot by the Dutch and IDA cofinanciers. Through energetic and resourceful IDA supervision, Koudougou Pilot started to take a coherent form late in its life, but it was too late. IDA had decided to shift its rural-development strategy to country-wide, functional projects. The Dutch, meanwhile, did not join IDA in this shift but have, after many years of hesitation, reformulated Koudougou. If this is a joint project, then it is surely curious to be auditing it while the Dutch part is still going on. C. Extension. 6.11 Was bringing T&V to Burkina Faso useful? Was the National System an improvement on it? Is the French critique well founded? 6.12 The fight between technological/ecological optimists and pessimists is reflected in disagreements over the best extension approach. During the earlier phase and, increasingly explicitly, during most of the middle phase under audit here, IDA was promoting the training & visit (T&V) system. Without any doubt, this initiall- made a great improvement by introducing order, norms and accountability. It also became increasingly clear a) that T&V's recurrent costs were unbearably high for an extremely poor country with meagre fiscal resources91 and b) that its elitist contact-farmer system was inappropriate to Burkinab6 rural society which exhibits a high degree of solidarity. 48/ "Koudougou Agricultural Development Project; Discussion on Reason for Late Withdrawal of Project by Bank," internal memo of 16 April 1981. 49/ For an assessment of the crippling recurrent-cost consequences of T&V extension in a far-away and not-so-poor country, see Stephan Seegers. The Training and Visit System and the Links between Rice Research and Extension in the Matara District of Sri Lanka. The Hague: International Service for National Agricultural Research, October 1990. It concludes (p. 1), "...after the World Bank stopped providing funds<,> The government ... was unable to assume the full costs of maintaining the agricultural research and technology transfer activities, and the number of research, extension, and training officers in the districts and regional centers decreased, as did the length and frequency of research-extension links." For more, see p. 9. OED's audit of five T&V extension projects in India (Report #8808 of 29 June 1990) deals with the unsustainability of T&V systems in West Bengal, Bihar, Kerala, Maharashtra and Tamil Nadu due to high recurrent cost and with the states' adaptations of the system to something they could live with. See Evaluation Summavy, p. xxiii and p. 35. - 24 - 6.13 The Borrower took the appropriate next step, adapting T&V to local conditions in the National System, which IDA (but not other cofinanciers) have accepted in the present phase of projects. Under the National System, it is recognised that extensionists may appropriately be used for data gathering and certain other administrative tasks. Instead of working with contact farmers, agents organize collective "field schools," usually on poor quality land, where they direct villagers in cultivation, and less successful "demonstration herds" for pastoralists. In this way, each agent works with a much larger group of clients than in classical T&V, so that unit costs are lower. The system's adoption appears to be successful, but the following problems should be kept in mind: a) field agents now consider themselves overworked and are nostalgic for the good old days of T&V; b) the system is just as top-down and rigid as T&V and agents soon run out of useful things to say. Attendance at groups, initially high, starts to decline.50/ Quality of attendees declines; and c) given the paucity of extension messages, the system is still too expensive. Other development agencies, notably the CCCE, argue that the National System reforms are an inadequate adaptation because they retain T&V characteristics that are faults under Burkinab6 conditions. In addition to the points mentioned above, they argue that the National System, like T&V, emphasizes structure over content. It is on the availability of appropriate extension messages that the CCCE critique of the National System manifests its technological/ecological pessimism. 6.14 The divergent assessments have led to harsh words and divergent interventions. IDA has approved country-wide projects in support of the National System and of agricultural research. The French are proposing to go back to the drawing board to find interventions that work by using mobile, multi-disciplinary teams that will listen to villagers and find SO/ Of course, while T&V is, perforce, somewhat rigid, is it not, in theory, supposed to be top-down. See D. Benor & J.Q. Harrison. Training and Visit Extension. Washington: World Bank, 1977 or D. Benor & M. Baxter. <same title & publisher>, 1984. In practice, this has not worked. To cite Seegers' conclusion from Sri Lanka (Op.cit. p. 8), "Host extension officers even had a top-down idea of the concept of information itselfs they only considered data from people higher in the hierarchy to be information." Seegers also found that "fewer and fewer questions were asked ... as time went on" ... "because they had been receiving the same message for a number of years." This description fits the Burkinabd situation perfectly. - 25 - better ways of farming and ranching.51/ The CCCE Tapoa project is an outgrowth of this conception, the unit costs of which will undoubtedly be very high. To bridge the gap, about 30 people from both sides52/ met in Burkina Faso in November 1989. In small groups, they toured the countryside for three days, then met for discuesion in Ouagadougou where IDA's Agricultural Director made a conciliatory presentation,53/ only to be sharply attacked. At the end of this "palaver under the tree," as one participant said, all saw the same rural Burkina Faso but IDA participants still wore rose-colored glasses and French still wore blue-colored glasses. The French, and other development agencies, politely declined the Borrower's invitation to cofinance PRESAP with IDA. The divergences are not healed, although the Village Land Management Project, now under preparation after joint appraisal by IDA, CCCE and GTZ, is a move in that direction. 6.15 The Audit agrees that the National System was several steps forward in adapting classical T&V to Burkinab6 conditions. The Audit finds that further adaptations would be desirable to make extension more responsive to its beneficiaries, and to increase flexibility so that, where there is little content to be extended, the formal structure can be changed and unwarranted public costs can be avoided.54/ The Audit does not, however, share the conviction of the more radical critics that current extension messages are worthless or environmentally damaging. The Audit believes that, when the historical process of population increase and intensification is understood, there are places where there are techniques worthwhile extending with current extension structures which the Borrower will, hopefully, continue to adapt to country conditions. This belief certainly does not detract from the priority of problem-oriented, client- sensitive research to address the manifest remaining problems -- how to make livestock and farming intensification intensification both sustainable and cheap enough to interest farmers and herders. 51/ The most comprehensive presentation of current CCCE thinking on the subject is Groupe de Travail Coop6ration Francaise. Les interventions en milieu rural: principes et approche m6thodolopique. Paris: Ministare de la Coop6ration et du Developpement, 1989. See also C.BaRRIER. Developpement rural en Afrigue de 1'Ouest soudano sahelienne: premier bilan our 1'approche zestion de terroir villageois. Paris: CCCE, mars 1990; and J.M.Yung & J.Zaslavski. Pour une prise en compte des strategies des producteurs: ragport provisoire. Montpellier: CIRAD & Paris: CCCE, sout 1990. 52/ The Burkinab6 were not invited. 13/ Based on the AGR work which has subsequently been published as Africultural Extension: The Next Step. World Bank. Policy and Research Series 1 13, 1990. 14/ As OED's audit of Indian T&V projects argues, "Investment in Intensive Extension Service may not be warranted in all circumstances." (Report P,808, p. 45). - 26 - D. Area-based or country-wide. 6.16 Which rural-development strategy is more appropriate to Burkinabd c, nditions at this time? 6.17 The CRPA staff the Audit worked with unanimously favored area- based agriculture projects,in which they have a vested interest, over country-wide ones. In fact, there have been start-up problems associated with the country-wide projects. It goes bnyond the problem of the more- than-two year delay in getting motorbikes with PRESAP financing through international competitive bidding. The severely strapped Government will have trouble financing its share of the CRPAs. Even though the National System is cheaper than T&V, CRPAs are bulky organizations and costly ones. Although their legal status suggests that they can be self-financing, this is clearly impossible. Cost recovery is almost never successful for extension services, which is mainly what CRPAs are. Their services from which cost recovery was possible were removed when they were re-organized from ORDs. In view of these problems and of Burkina Faso's relatively favorable experience with area-development projects, wouldn't it be wiser to keep them decentralized, appealing to foreign financial sources for help when necessary? 6.18 The Audit assumes that that logistical and procurement problems can be mastered. If so, the Audit believes that country-wide projects are suitable approaches to development of rural-support activities such as extension, research, even credit when the circumstances justify it. However, farming and animal husbandry are extremely site-specific activities, less adapted than any other to centralized planning and control. Unless the country-wide programs are decentralized and adapted to ensure that interventions in technology and services are appropriate for conditions in specific areas, they will not be cost-effective. By their nature, geograp..cally-based projects have the inherent opportunity to provide interventions to change farming, livestock raising, and land management practices in the contexy of a particular area. The high cost of these projects, however, limits their replicability. Nevertheless, this should not preclude their use in particular situations, e.g., where critical environmenta:., social or economic problems in an area cannot readily be handled by a ccuntry-wide program and external funding is available to support a development solution through its full implementation; if the project does not involve an unjustified amount of Government counterpart funding, nor divert qualified national staff from higher priority programs, nor leave Government with an unwanted current expenditure after completion, then the area-development project would logically have legitimacy. Particular circumstances could even justify Government funding of an area-development project alongside its major country-wide projects if the problems to be addressed in an area are important enough and cannot be effectively resolved through its main programs. - 27 - E. Rural credit. 6.19 What improvements are desirable in the present situation, in which large parts of the four-projects' area don't have any? 6.20 Regarding rural credit, the Audit finds the present situation anomalous. Credit was a key facet of the projects under audit. As the task manager for Koudougou Pilot said, "Credit is at least as important as technical aspect., since the farmer, in order to apply the new methods, must have access from the beginning to animals, equipment and the various inputs required. (All of these are interrelated: farmers will not adopt improve animal nutrition practices if their animals are idle for lack of equipment.) Because the initial investment is crucial, the credit aspect cannot be handled separately."55/ Today, in major portions of the four-projects' area, investment credit cannot be had at all, either from the informal sector or from caisses populaires or from the National Agricultural Credit Bank (CNCA) through the village groups (GV). CNCA has wisely refrained from expanding more rapidly than its human and organizational resources make prudent. But that fact leaves large parts of rural Burkina Paso effectively uncovered by the credit network. 6.21 The technicians who, until recently, implemented the four projects -- interested parties to be sure -- feel strongly that they should get their old role in running the official credit system back. They argue that extension and credit are inexorably intertwined and, further, that their extension messages do not sell themselves, so that their extension effectiveness is reduced when they have no role in credit. 6.22 The Audit is mindful of IDA doctrine, which antedates T&V extension, that confidence in extension workers is destroyed when they double as moneylenders, most specifically when they have to dun clients to pay up. The Audit is also impressed by the argument of the extension workers of the CRPAs in question. It is, regrettably, beyond the scope of this audit to decide which is right. We will go so far as to say, however, that in rural Burkina Faso's present circumstances in which neither CNCA nor the caisses populaires have sufficient articulation to serve most of the country, some sort of credit role for the CRPAs would be good for development. Around the world, imprudent and/or corrupt credit components Inave been the ruin of many rural development projects. A possible way of having the benefits of CRPA involvement in credit without too many of the social costs would be to have CREAs provide credit services to CNCA and possibly others on a fee-for-services basis. This would represent an extension of the arrangements which have grown up between some CRPAs and CNCA for collection of outstanding GV debts. Any credit administration through CRPAs would have to be organized to minimize any imposition on the 55/ Internal memo of 10 January 1986, Pare. 6. - 28 - primary extension function of field staff, possibly through a minimal number of specialized supporting personnel. 6.23 The Audit regrets that the above findings have a definite tentative quality. That is dictated, however, by the facts at our disposal. 6.24 It is disappointing, one to two years after credit disbursements are complete, not to know more than we do about the impact and about several critical issues concerning these projects. The Audit feels that its conclusions about issue A -- ecological effects -- are robust. That confidence, however, is not based so much on data generated by the projects and by their M&E units as on independent infermation about the milieux. With respect to issue C -- extension -- it has to be disappointing not to have confidence about the magnitude of social benefits that considerable public expenditures on this service are producing. Most troubling is the conclusion of the CCCE audit of Boucle du Mouhoun and Haute Bassins discussed in Para. 5.05 above and endorsed by the Audit. After 9 to 10 years, we cannot be sure whether production increases claimed for the projects stemed from them or not. The audit, however, appreciates the difficulties inherent in measuring the impact of extension projects. 6.25 The Audit has commented on the inadequacies of M&E of the Koudougou Pilot (principally Para. 4.08, but see also Paras. 4.02 and 4.09- 4.10). Since Koudougou was supposed to be a pilot project preparing the way for further investments, good M&E was particularly pertinent. More to the point, its failure to provide pertinent information weakened the case for a sequel. The Audit could also have discussed salient M&E weaknessec in Boucle du Mouhoun, discussed in the PCR, and in Bougouriba II, barely alluded to in the PCR but quite evident at Audit. The expenditure on expatriate expertise in all four projects did not solve the problem and may even have been part of it in one or two cases. 6.26 Disappointing experience with M&B is certainly not a unique feature of these four projects. It is the rule, rather than the exception. In these cases, as elsewhere, M&E units suffered from competition between contrary objectives: answering the immediate questions of day-to-day management and the deep questions, such as effects on soil fertility. They suffered for the contradictory expectations of their two audiences: their project directors, interested in daily management, and the financiers' supervisors, interested in keeping disbursevents on schedule. Both were preoccupied with issues other than those that this Audit, with its luxury of historical perspective, has raised. 6.27 Nevertheless, the M&E disappointment in this case is noteworthy, and not merely because some foreign experts proved to be inappropriate. For a very poor country like Burkina Paso with meagre sources of public funds, programmes like T&V extension, even when modified as in the National System, and the other aspects of the CRPAs are quite expensive. It is quite galling not to be surer that the money was well spent. This raises the serious question as to what can external financiers really expect from M&E components in development projects. In some cases the techniques to reliably and statistically measure impact are not readily available, e.g., - 29 - in extension and research projects. In most cases the incentive to measure impact lies more with the donor or financier than with the implementing agency. After all, the Borrower, and especially the executing agencies, would not normally undertake a project unless they considered it to be worthwhilel their primary objective then logically becomes project implementation, not statistical baseline studies or comparisons with control treatments to give data on "with and without" situations. Perhaps the financiers and donors have to consider covering the cost of impact monitoring and evaluation in at least a sample of project types to provide the hard data required to assess development approaches. - 31 - ATTACHMENT Page 1 of 4 COMMENTS FROM COFINANCIER TRANSLATED VERSION OF THE TELEX FROM CAISSE CENTRALK DE COOPERATION ECONOMIQUE SENDER J.C. DEVEZE Service reference: DPE/DRS FAX # CALLED: 202.477.6391 Paris, 13 May 1991 ADDRESSEE: World Bank/ Washington Mr. Graham DONALDSON Chief, Agriculture, Infrastructure & Human Resources Division Operations' Evaluation Department SUBJECT/OBSERVATIONS: Please find attached comments on the preliminary version of the evaluation report of Volta Noire and Haute Bassins projects. We wish you good reception of them. NUMBER OF PAGES 3 Signature: R.HIRSCH Chief, Retrospective Evaluation & Sectoral Studies Division S32 - ATTACHMENT Page 2 of 4 CAISSE CENTRALE DE COOPERATION ECONOMIQUE Public Enterprise - Capital 1 billion France Paris, 13 May 1991 Mr. Graham DONALDSON Chief, Agriculture, Infrastructure & Human Resources Division Operations' Evaluation Department World Bank 1818 H Street DC 20433 WASHINGTON (USA) ref: Your letter of 26 March 1991 N/REF.DPEIDRS/JCDICA Dear Sir, The CCCE's services have come to know the preliminary version of the retrospective evaluation report on the different Burkinabe rural development projects with great interest. The step of presenting in a single document the evaluation of several projects in a single country and in a single sector seems quite apt for it brings about a confrontation of the methods, strategies and results of the projects, which facilitates a sectoral approach. The evaluators have correctly emphasized the difficulty of discerning which increases in agriculturol production were due to the projects and which were related to the context (migrations, price changes, etc.). One must not forget that the interventions analysed form part of the dynamic of cotton development, which has depended on the security that the organization of that line of production has brought to producers. More specifically, as regards the Volta Noire and Hauts Bassins projects, it seems to us that it would have been desirable to better explain the reasons for the important amounts (62%) of the credit cancelled for these two projects (see p. 1 of preliminary version). In this regard, it is interesting to note that certain components, such as project management and extension, were better implemented in terms of expenditures than were supporting research, support to herders who have settled down, and projects to be approved (no doubt because the agricultural services are more given to making things already in place work than to preparing and sustaining innovations). -33 - ATTACHMENT Page 3 of 4 As far as the assessment of the returns to the projects is concerned, the evaluators present the very high rates of return (30+%) for the two projects which were calculated by the projects' monitoring and evaluation units. It certainly seems that these figures are nothing but very summary estimates (and probably very optimistic). It is worth stressing this point, because that would avoid ambiguities in interpretation. The debate on extension approaches, which is opened in the report, is more fundamental. Thus, as we wrote you already regarding the project completion report for these projects (letter of 13 March 1991), "from our perspective, the principal lesson is that putting in place heavy structures for intervention to increase agricultural production by * introduction of new techniques probably doesn't respond to the complexity of situations. Another way, which depends more on rural populations as responsible partners involved in the elaboration of new solutions, seems desirable for installing a long-term, sustainable development process. But this supposes that the methods of intervening in rural areas1/ and the corresponding human means that need to be mobilized be reviewed. Thus it appears to us that the way the evaluators attribute to IDA a favorable role in improving the extension component (p. 7) and in instituting a new extension system (pp. 18 & 23) amounts more to affirmations of principle than to a supported demonstration. There is a certain contradiction between the affirmation that the extension system has been adapted to the local situation (p. 31) and the observation that problems persist (p. 32). One might ask whether the Borrower, in choosing a national extension system, was not influenced both by the possibility of obtaining important financin, in order to cope with the expense of the existing staff, and also by the temptation to reinforce the centralization of the system. Our differences in assessment also stem partly from different ways in which the questions posed are understood. The IDA evaluators ask (pp. 1 & 2) %;-hether, insofar as the results in terms of production were deemed satisfactory, localized projects (also called integrated projects) should be pursued or not, while the CCCE is debating the possibilities of pursuing development interventions with the concerned populations, taking into account lessons learned following the difficulties encountered. A propos, the opposition between localized agricultural projects and centralized programs (see p. 34) seems to be a false debate, the important thing being to articulace interventions between national, regional, anA local levels while taking the context into account. The way in which the rural milieu perceived these projects appeared definitive at the time of CCCE's retrospective evaluation because it - 34 - ATTACHMENT Page 4 of 4 makes possible a reflection on the pertinence of the rural development strategies adopted. Let's remember on this point IBRD's retrospective evaluation of the first cotton project which stressed, even then, the "academic" character of the notion of "average farms." In spite of all subsequent efforts, the failure to take into account the specificities of the rural world in the Volta Noire and Hauts Bassins projects and, especially, of migration,2/ which was known and had been analyzed in detail, this failure was probably responsible in good measure for the meagre rural development results. Finally, it is too bad that the evaluators didn't go deeper in examining the projects' results in peasant organization because this aspect is important for increasing farmer participation in development activities. It was, however, one of the points made in the conclusions of Cotton Development Programs in Burkina Faso, C&te d'Ivoire, and Togo (World Bank Operations' Evaluation Staff, 1988). Sincerely yours, For the Director of the Policies and Studies Department Jean-Claude DEVEZE P.S. The CCCE did a retrospective evaluation (in addition to a completion report), which calls for a correction in the Preface (p.M1). 1/ The projects have not promoted an intervention method simultaneously taking into account the "enterprise" needs (advice on agriculture and management), the evolution of village groups (support for managing them), and what is necessary for management of village lands. 2/ "Differences" between persons or institutions cannot be an explanation of the failure to take migration into account when the projects were conceptualized. IBRD 15915R BURKINA FASO ~ ) HAUTS-BASSINS AGRICULTURAL DEVELOPMENT PROJECT _PROET DE DEVELOPPEMENT AGRICOLE DES HAUTS-BASSINS ,* ORD Supply Centers Mogosn de /'CRI OBANGASSO Veternary Posts Poste Vétérnare OMOROL ABA Ginneres Usine d'dgren<ge AVV Zone Zone AVV Apiphed Research Unit ?IjQNKÉLÉ -o ou Un'té de Riecherche d'Accompgnement N'DOROL A Seed Form lefmo Jemencib .ÉKOUNDOUGOU SAFrmRAmncd O SILORL A ,0NIENA *,,BOSSORA Forestry Prolect Area Zone Prot Porestier Nieno- Donkele Prouect Area EREBA Zone noter Mén, - Dionke/e KOUROUMA Mo fo:rest Settled Pastoraists Areas DAND s BONI Zone /ev~urs Sdentonses - ~SÉGUÉRÉ d Q, OBOUAHOUN ^HO D ORD Headquarters ___ Chef- /,e de I' D Sector Offices Chef- beu de Secrer Sector Boundones Lmite de Secrier 11 SA ORD Boundares limite de /'CWD 100 Intemational Boundanes From1br,e National Roads KOUMBIA f~ Nøio OKARANKASSO Socondary Road Route Seconidi -D....-E---Railway Chemin de Før Rivers co,,s d'ecu lsohyets in Millieers OLOKO 7s0yers en M/hmetres AHON OKOTOURA ORODARAN GER . BURKINA FASO TOUSS91AÑ,. .j. Agatultro Developrnent 0 Ouogodoeouge - - ~. -,.. Pro1ect Area Bank ond the loterneahinl F,ane C.poraion. The den nintions sed nd ME TERS G H A N the bemndanes shown en this nop do notiimply, en the port ef The World Bank c' ond the 1t~ernetienel Finence Cerperetion, any jd ft on the 1eg9& status e& Hats y ony eor any e nt r apteo ch b r COTE D IVOIRE JULY 1990 IBRO 】5127R】 -,（긔 IBRD 14815R MALI NIGER BURKINA IAS0 Z'> < YAK > (BENIN GHANA i i ARBOLLE jTOG0 0 BURKINA FASO KOUDOUGOU AGRICULTURAL OG DEVELOPMENT PROJECT k SAMBA- lo..N0R0 P5f0JETOF DEVELOPPEMEN7' AGPICOIE DE K0UDOLIG0U Kw 1967'-19771S0hyets D Agriculturol zones IDYIR de zom ogri.cole ORD/d partment boundaries Zimitede4ýefportemont,'ORD 12-30- Sector boundories KIND10ý11ý- L imite de seeteur --------- Sub-sector boundories Lmite de sous-sectrur Deporment capital \oR'E chel-l'æ de dý#portsment sector capitois 0 YAKO chef-lieu de seeleur K6'DOILIGOU OSaI-- 0 SAMBA Sub-sector copitals sous-sectew KOKOLOGO 0 so,.. Ofher villages TEINADO,' Autres vdlages ro ..90. Rodwoy chemin d& fer Main roods Routes principale$ ------- ----- Ma n trocks SABOU e$ pincipoles irco, Nik. R.ve,s POUN I cours dýau 12,0cr 0 1 Internotiono. boundary Frontiere KILOMETERS 0 3 10 15 2,0 2,5 3,0 35 -X L -0 IPKASSOU RFARA SAPOVY 11-30' NABOU TIIbo. it Yh" ..p h.. h- p...p-d b, Wdd Bod,. .P -d Ø-k-1 Wrld 8.4 -d f. it h.., ..p d. p- . Th. i. 0 b-d-- LEO -B. No ý~ G H A N A JULY 1990 &：「나 万 JUNE 1 99】
Groupe de la Banque mondiale · Project Performance Assessment Report
Upper Volta - Agricultural Development Projects
Voir le document original
Le texte intégral est hébergé par l’organisation qui le publie. lawenc.com indexe les métadonnées et renvoie vers la source officielle.
Texte intégral
Informations clés
Organisation
Groupe de la Banque mondiale
Type de document
Project Performance Assessment Report
Pays
Burkina Faso
Source
Banque mondiale